An electric vehicle intelligent vehicle lamp adjusting system, method, computer device and storage medium

Abstract

This application provides an intelligent headlight adjustment system, method, computer device, and storage medium for electric vehicles. The sensor module acquires the driver's heart rate, road surface ultrasonic reflection signals, road surface images, and vehicle steering and tilt angles. The data processing module determines changes in heart rate to generate a first instruction to adjust the light intensity range; it extracts features from ultrasonic reflection signals and road surface images to identify road surface materials and generates a second instruction to adjust the light intensity; it fuses vehicle speed and lateral acceleration to determine the turning state. If turning is detected, it triggers high-frequency operation of relevant detection units to process angle data and calculate adjustment values ​​based on fixed illumination area parameters, generating a third instruction to adjust the light angle and range. The control execution module drives the headlight module to adjust the light angle, range, and intensity according to these instructions, thus achieving headlight adjustment. This solution ensures accurate light adjustment in various scenarios, significantly improving the safety of electric two-wheeled vehicles at night and in complex road conditions.

Description

Technical Field

[0001] This invention relates to the field of vehicle automatic control, and more specifically, to an intelligent headlight adjustment system, method, computer device, and storage medium for electric vehicles. Background Technology

[0002] With the increasing popularity of electric two-wheelers in short-distance urban travel, driving safety at night and in complex road conditions has become a growing concern for the industry and users. As the core lighting component of a vehicle, the rationality of the adjustment of vehicle lights directly determines the clarity of the driver's vision, thereby affecting the effectiveness of driving risk prevention and control.

[0003] Most electric two-wheelers on the market use fixed parameters for their headlight systems. The lighting angle, range, and intensity are preset at the factory, and in actual use, they only support manual switching between "high beam" and "low beam," making the adjustment method relatively simple.

[0004] However, the study found that the above-mentioned solutions have obvious scene adaptation defects and cannot dynamically adjust the lighting parameters according to the actual driving scenario. For example, when driving straight at night, the fixed illumination range is difficult to cover sudden lateral obstacles, and the driver may miss the opportunity to avoid danger due to limited vision. When facing different road surfaces such as asphalt and gravel, even if the headlights output the same light intensity, the effective light reflected to the driver's field of vision is significantly different due to the difference in the light absorption rate of the road surface. This can easily lead to the driver's delayed recognition of road potholes and debris, indirectly increasing the safety hazards of nighttime driving. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an intelligent headlight adjustment system, method, computer device and storage medium for electric vehicles, so as to ensure accurate lighting adjustment in various scenarios and significantly improve the driving safety of electric two-wheelers at night and in complex road conditions.

[0006] In a first aspect, embodiments of this application provide an intelligent headlight adjustment system for an electric vehicle, the system comprising a sensor module, a data processing module, a control execution module, and a headlight module; The sensor module is used to acquire the driver's heart rate data, ultrasonic reflection signals from the road surface, road image data, and vehicle handlebar steering angle data and vehicle body tilt angle data. The data processing module is used to determine the heart rate change characteristics based on the heart rate data, and generate a first light adjustment command for adjusting the light intensity and light range based on the heart rate change characteristics. The data processing module is further configured to determine the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, respectively, identify the road surface material based on the road surface image texture features and ultrasonic reflection features, and generate a second light adjustment command for adjusting the light intensity based on the road surface material and the light absorption rate of different materials. The data processing module is also used to fuse the vehicle speed signal and the lateral acceleration data of the vehicle body, and use a decision tree algorithm to determine whether the vehicle has entered a turning state; if it is determined to be turning, the steering detection unit and the tilt detection unit are triggered to switch from low power standby mode to high frequency working mode, receive and process steering angle data and tilt angle data, calculate the illumination angle adjustment value and illumination range expansion value in combination with preset fixed illumination area parameters, and generate a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value; The control execution module is used to drive the vehicle headlight module to adjust the illumination angle, illumination range and illumination intensity according to each illumination adjustment command, so as to realize the vehicle headlight adjustment.

[0007] Optionally, the sensor module includes a heartbeat detection unit, a road surface detection unit, a steering detection unit, and a tilt detection unit; The heartbeat detection unit is installed in the area where the handlebars contact the palm, and is used to continuously collect the driver's heart rate data at a preset sampling frequency, and transmit the heart rate data to the data processing module in real time. The road surface detection unit includes an ultrasonic sensor and a camera sensor. The ultrasonic sensor is used to collect ultrasonic reflection signals from the road surface at a preset sampling frequency. The camera sensor is used to collect road surface image data at a preset frame rate. The ultrasonic sensor and the camera sensor achieve timestamp alignment of data acquisition through a synchronous trigger signal, and transmit the ultrasonic reflection signals and the road surface image data to the data processing module. The steering detection unit is installed at the rotation axis of the handlebars and is a Hall effect angle sensor with a preset angle measurement range and measurement accuracy. The tilt detection unit is installed near the vehicle's center of gravity. It is a MEMS inertial measurement unit that integrates a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer. It is used to measure the vehicle's tilt angle and attitude, and has a preset attitude measurement accuracy. The steering detection unit and the tilt detection unit are in a low-power standby mode when the vehicle is not turning. When the vehicle enters a turning state, they are triggered to start, and collect the handlebar steering angle data and the vehicle body tilt angle data respectively and transmit them to the data processing module.

[0008] Optionally, determining the heart rate change characteristics based on the heart rate data and generating a first light adjustment command for adjusting light intensity and light range based on the heart rate change characteristics includes: The heart rate data is filtered in real time to remove noise interference and obtain a heart rate numerical sequence; Heart rate change characteristics are calculated based on the heart rate numerical sequence, wherein the heart rate change characteristics include instantaneous peak heart rate, duration of heart rate exceeding threshold, and rate of heart rate rise. Preset heart rate threshold range and corresponding light adjustment rules: When the instantaneous peak heart rate exceeds the upper limit of the heart rate threshold range, or the heart rate continues to exceed the threshold for a preset duration, or the heart rate rise rate exceeds the preset rate, the heart rate change characteristics are determined to meet the adjustment conditions. The first illumination adjustment command is generated based on the satisfied adjustment conditions: the greater the heart rate exceeds the threshold, the longer the duration of exceeding the threshold, or the faster the heart rate rises, the higher the increase ratio of light intensity and the greater the expansion angle of the illumination range in the first illumination adjustment command.

[0009] Optionally, the step of determining the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data respectively, identifying the road surface material based on the road surface image texture features and ultrasonic reflection features, and generating a second illumination adjustment command for adjusting the illumination intensity based on the road surface material and the light absorption rate of different materials includes: The ultrasonic reflected signal is preprocessed by amplifying the weak signal and filtering out noise with a bandpass filter to extract ultrasonic reflection features, which include the reflected signal amplitude ratio, propagation time deviation and pulse width variation coefficient. The road surface image data is preprocessed by Gaussian filtering algorithm for noise reduction and histogram equalization algorithm for contrast enhancement. Convolutional neural network algorithm is used to extract the texture features of the road surface image. The texture features of the road surface image include local binary mode (LBP) histogram features and contrast and correlation features of gray-level co-occurrence matrix (GLCM). Construct a road surface material feature database, which stores ultrasonic reflection feature samples and road surface image texture feature samples corresponding to different road surface materials; The extracted ultrasonic reflection features and road surface image texture features are matched with samples in the road surface material feature database. The material corresponding to the sample with the highest matching degree is selected as the identified road surface material, and the light absorption rate parameter corresponding to the road surface material is retrieved. The expected road surface illumination intensity required for vehicle operation is preset. Based on the light absorption rate parameter corresponding to the road surface material and the expected road surface illumination intensity, the target emitted light intensity to be output by the headlight module is determined, and the second illumination adjustment command is generated.

[0010] Optionally, the fusion of vehicle speed signal and lateral acceleration data, using a decision tree algorithm to determine whether the vehicle has entered a turning state, includes: The vehicle speed signal and lateral acceleration data during the vehicle's movement are acquired, and the two types of data are processed for time synchronization and noise filtering. Construct a judgment feature system for the decision tree algorithm. The feature system includes key features related to vehicle turning behavior and corresponding judgment thresholds. The key features include at least vehicle speed features, vehicle lateral motion features, and feature state duration. Based on the aforementioned judgment feature system, a progressive decision tree judgment logic is established: first, the vehicle speed feature is used to determine whether the vehicle is in a speed range that is easy to turn; then, the vehicle body lateral movement feature is used to determine whether the vehicle body generates lateral movement related to turning; finally, the duration of the feature state is used to determine whether the above features constitute a stable turning-related state. The decision tree algorithm is used to comprehensively analyze the judgment results of the above features and output a judgment on whether the vehicle has entered a turning state. Based on the judgment result, the corresponding operation is triggered: if it is determined that the vehicle has entered a turning state, a trigger signal is generated to control the steering detection unit and tilt detection unit to switch from low power standby mode to high frequency working mode; if it is determined that the vehicle is not in a turning state, the low power standby mode of the steering detection unit and tilt detection unit is maintained.

[0011] Optionally, the step of calculating the illumination angle adjustment value and the illumination range expansion value based on preset fixed illumination area parameters, and generating a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value, includes: Preset fixed lighting area parameters, wherein the fixed lighting area parameters include the ideal area length L0, width W0, and headlight installation height h of the area illuminated by the headlights when the vehicle is traveling straight; Receive and process the handlebar steering angle data θ collected by the steering detection unit and the vehicle body tilt angle data γ collected by the tilt detection unit; The illumination angle adjustment value is calculated based on the principle of geometric optics, including the horizontal turning angle adjustment value α=θ / 2 and the vertical tilt angle adjustment value δ=arctan(h / (L0cosγ)); The illumination range extension value β=θ is determined based on the handlebar steering angle θ, and the illumination range is extended towards the steering side; The horizontal steering angle adjustment value α, the vertical tilt angle adjustment value δ, and the illumination range expansion value β are written into the third illumination adjustment command, and the direction, magnitude, and range expansion direction and angle of the headlight angle adjustment are specified.

[0012] Optionally, the step of driving the headlight module to adjust the illumination angle, illumination range, and illumination intensity according to each illumination adjustment command to achieve headlight adjustment includes: The control execution module is configured with an angle range adjustment unit and an intensity adjustment unit. The angle range adjustment unit is used to drive the vehicle lamp module to adjust the illumination angle and illumination range, and the intensity adjustment unit is used to adjust the illumination intensity of the vehicle lamp through current control. When the first illumination adjustment command is received, the angle range adjustment unit drives the headlight module to expand the illumination coverage range, and the intensity adjustment unit improves the illumination intensity of the headlight by smoothly adjusting the driving current of the headlight source. When the second illumination adjustment command is received, the intensity adjustment unit calculates the difference between the current illumination intensity and the target illumination intensity based on the target emitted illumination intensity determined in the second illumination adjustment command, and gradually adjusts the illumination intensity to the target value through a closed-loop control strategy. When the third illumination adjustment command is received, the angle range adjustment unit drives the headlight module to adjust the horizontal and vertical illumination angles, and at the same time drives the headlight module to expand the illumination range towards the turning side; if the angle and range adjustment causes a decrease in the illumination intensity per unit area, the intensity adjustment unit synchronously compensates and adjusts the illumination intensity. When multiple illumination adjustment commands are triggered simultaneously, the control execution module responds first to the third illumination adjustment command to complete the adjustment of the illumination angle and range; then, based on the adjusted angle and range, it superimposes the illumination intensity adjustment requirements from other illumination adjustment commands.

[0013] Secondly, embodiments of this application provide an intelligent headlight adjustment method for electric vehicles, applied to an intelligent headlight adjustment system for electric vehicles. The system includes a sensor module, a data processing module, a control execution module, and a headlight module. The method includes: The sensor module acquires the driver's heart rate data, ultrasonic reflection signals from the road surface, road image data, and vehicle handlebar steering angle data and vehicle body tilt angle data. The data processing module determines the heart rate change characteristics based on the heart rate data, and generates a first light adjustment command for adjusting the light intensity and light range based on the heart rate change characteristics. The data processing module determines the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, identifies the road surface material based on the road surface image texture features and ultrasonic reflection features, and generates a second light adjustment command for adjusting the light intensity based on the road surface material and the light absorption rate of different materials. The data processing module integrates the vehicle speed signal and the vehicle's lateral acceleration data, and uses a decision tree algorithm to determine whether the vehicle has entered a turning state. If it is determined to be turning, the steering detection unit and the tilt detection unit are triggered to switch from low-power standby mode to high-frequency working mode, receive and process steering angle data and tilt angle data, and calculate the illumination angle adjustment value and illumination range expansion value in combination with preset fixed illumination area parameters. Based on the illumination angle adjustment value and the illumination range expansion value, a third illumination adjustment command is generated to adjust the illumination angle and illumination range. The control execution module drives the vehicle headlight module to adjust the illumination angle, illumination range and illumination intensity according to each illumination adjustment command, so as to realize the vehicle headlight adjustment.

[0014] Thirdly, embodiments of this application provide a computer device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the electric vehicle intelligent headlight adjustment method described in any of the optional embodiments of the second aspect above are performed.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the electric vehicle intelligent headlight adjustment method described in any of the optional embodiments of the second aspect above.

[0016] The technical solution provided in this application includes, but is not limited to, the following beneficial effects: The sensor module acquires driver heart rate data, ultrasonic reflection signals and image data from the road surface, as well as vehicle steering and tilt angle data, enabling comprehensive perception of multi-dimensional information about the "human-vehicle-environment". This multi-source data acquisition method provides rich and accurate raw data for subsequent headlight adjustments, breaking the limitations of traditional headlights that rely on a single parameter or manual adjustment. It allows headlight adjustments to dynamically adapt to the actual scenario, ensuring the targetedness and effectiveness of the adjustments from the source.

[0017] The data processing module generates the first illumination adjustment command based on heart rate data, linking the driver's physiological stress state with the headlight adjustment. When the driver experiences a sudden increase in heart rate, a sustained exceedance of the threshold, or a rapid increase due to complex road conditions, the system can promptly increase the light intensity and expand the illumination range to provide the driver with clearer vision support, helping them better cope with potential risks. This achieves real-time linkage between headlight adjustment and driver status, improving driving safety in emergency scenarios.

[0018] The data processing module identifies road surface materials based on ultrasonic reflection characteristics and road surface image texture features, and generates a second illumination adjustment command by combining the material's light absorption rate. This solves the problem of uneven lighting effects caused by differences in light absorption among different road surface materials. By dynamically adjusting the light intensity, it ensures that different road surface materials such as asphalt, gravel, and soil can all receive preset effective illumination, avoiding delays in road surface detail recognition caused by material differences and guaranteeing lighting stability under various road surface conditions.

[0019] The data processing module determines the turning state by fusing vehicle speed and lateral acceleration, and then generates a third lighting adjustment command to specifically optimize lighting for turning scenarios. By adjusting the lighting angle, it compensates for the field of vision shift caused by steering and tilting, while expanding the lighting range on the turning side, effectively eliminating blind spots during turns. Combined with calculations of fixed lighting area parameters, it ensures that the lighting coverage during turns is consistent with that during straight-line driving, improving visibility and driving safety during turns.

[0020] The control execution module drives the headlight module to adjust the illumination parameters according to various illumination adjustment commands. When multiple commands are triggered simultaneously, it prioritizes the response to the third command related to turning, and then superimposes other intensity adjustment requirements. This collaborative execution mechanism not only ensures priority adaptation of lighting in key scenarios (such as turning), but also achieves comprehensive optimization of illumination parameters in multiple scenarios. This enables the headlights to provide optimal lighting effects under various complex driving conditions, comprehensively improving the safety of electric two-wheeled vehicles at night and on complex road conditions.

[0021] In summary, this invention achieves intelligent dynamic adjustment of electric vehicle lights through multi-dimensional perception, precise decision-making in different scenarios, and collaborative execution. Its beneficial effects include: deeply binding driver status, road environment, and vehicle posture with headlight parameters, enabling the lighting effect to adapt to different driving scenarios in real time. This effectively eliminates the limitations of field of vision and uneven lighting issues associated with fixed-parameter headlights, significantly improving the driving safety and comfort of electric two-wheelers in scenarios such as nighttime driving, complex road conditions, and turning.

[0022] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This diagram illustrates the structure of an intelligent headlight adjustment system for electric vehicles according to Embodiment 1 of the present invention. Figure 2 This shows a schematic diagram of the structure of the second type of intelligent headlight adjustment system for electric vehicles provided in Embodiment 1 of the present invention; Figure 3 The flowchart of an intelligent headlight adjustment method for an electric vehicle provided in Embodiment 2 of the present invention is shown; Figure 4 A schematic diagram of the structure of a computer device provided in Embodiment 3 of the present invention is shown. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0026] Example 1 To facilitate understanding of this application, the following is combined with... Figure 1 The schematic diagram of the intelligent headlight adjustment system for electric vehicles shown in Embodiment 1 of the present invention describes the contents of Embodiment 1 in detail.

[0027] See Figure 1 As shown, Figure 1 The diagram shows a schematic of an intelligent headlight adjustment system for an electric vehicle provided in Embodiment 1 of the present invention. The system includes a sensor module 101, a data processing module 102, a control execution module 103, and a headlight module 104.

[0028] The sensor module is used to acquire the driver's heart rate data, ultrasonic reflection signals from the road surface, road image data, and the vehicle's handlebar steering angle and body tilt angle data.

[0029] Specifically, the sensor module, acting as the system's "perception layer," undertakes the crucial task of multi-dimensional data acquisition. The data it collects covers three core dimensions: driver, road surface, and vehicle, providing a comprehensive and reliable data source for subsequent adjustments. Regarding driver physiological state data acquisition, an optical heart rate sensor (heartbeat detection unit) installed inside the driver's grip acquires real-time heart rate data. This data indirectly reflects the driver's stress response to the driving environment; for example, heart rate fluctuations are noticeable when encountering sudden obstacles or changes in road conditions, providing a basis for the system to assess environmental complexity. Road surface environment data acquisition relies on a road surface detection unit composed of ultrasonic and camera sensors: the ultrasonic sensor emits and receives ultrasonic signals reflected from the road surface at a preset frequency, capturing the physical reflection characteristics of the road surface; the camera sensor simultaneously captures road surface images, recording the visual texture features of the road surface. The two sensors achieve timestamp alignment through synchronous trigger signals, avoiding data mismatch caused by acquisition time differences, providing dual-dimensional support for subsequent road surface material identification. Vehicle attitude data acquisition is accomplished by a steering detection unit and a tilt detection unit. The former, mounted at the handlebar pivot, uses a Hall effect angle sensor to collect the steering angle, reflecting the vehicle's steering direction and amplitude. The latter, mounted near the vehicle's center of gravity, uses a MEMS inertial measurement unit to collect the vehicle's tilt angle, reflecting the vehicle's roll state during cornering. To optimize energy consumption, both units default to a low-power standby mode, switching to a high-frequency operating mode only when the system determines that the vehicle has entered a cornering state. This ensures timely data acquisition while avoiding power waste caused by continuous operation.

[0030] The data processing module is used to determine the heart rate change characteristics based on the heart rate data, and generate a first light adjustment command based on the heart rate change characteristics to adjust the light intensity and light range.

[0031] Specifically, the data processing module is the system's "decision-making center." It analyzes and processes different types of data collected by sensors, generating lighting adjustment instructions adapted to different scenarios. The first lighting adjustment instruction focuses on driver stress scenarios, with the core being the correlation between heart rate changes and lighting needs. First, the raw heart rate data is filtered using a combination of sliding window filtering and Kalman filtering algorithms to remove interference from hand tremors, sensor noise, and other factors, resulting in a stable heart rate sequence. Then, based on the sequence, three key features are calculated: instantaneous peak heart rate (the highest heart rate at a given moment), duration of heart rate exceeding the threshold (the cumulative time the heart rate exceeds the normal range), and heart rate rise rate (the speed at which the heart rate changes from normal to exceeding the threshold). These features quantify the driver's physiological stress level. The system presets a normal heart rate threshold range for drivers (e.g., 60-100 beats / minute) and corresponding adjustment rules. When any feature meets the "threshold exceeding" condition (e.g., instantaneous peak exceeding the upper limit, duration of continuous exceeding the threshold exceeding the standard, or excessively rapid rise rate), it is determined that the current environment may pose a risk, requiring enhanced lighting. During the instruction generation phase, the adjustment parameters are determined based on the "degree" of the heart rate exceeding the threshold: the greater the magnitude of the heart rate exceeding the threshold, the longer the duration, or the faster the rate of increase, the higher the increase in light intensity and the greater the angle of expansion of the light range. For example, when the heart rate exceeds the threshold by 20%, the intensity is increased by 40% and the range is expanded by 25°, ensuring that the lighting intensity dynamically increases with the complexity of the environment, assisting the driver in dealing with potential risks.

[0032] The data processing module is further configured to determine the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, respectively; identify the road surface material based on the road surface image texture features and the ultrasonic reflection features; and generate a second illumination adjustment command for adjusting the illumination intensity based on the road surface material and the light absorption rate of different materials.

[0033] Specifically, addressing the differences in road surface environments, the data processing module collaboratively analyzes ultrasonic reflection signals and road surface image data to generate a second illumination adjustment command adapted to the road surface material, resolving the uneven lighting problem caused by differences in light absorption rates among different road surface materials. In the dual data feature extraction stage, the ultrasonic reflection signal is first amplified by an amplification circuit, then clutter is filtered out using a bandpass filter circuit, ultimately extracting three types of features: the amplitude ratio of the reflection signal, propagation time deviation, and pulse width variation coefficient—these features respectively reflect the road surface's reflectivity, smoothness, and roughness. For the road surface image data, Gaussian filtering and histogram equalization are first applied to enhance contrast, then a convolutional neural network algorithm is used to extract Local Binary Pattern (LBP) histogram features and Gray-Level Co-occurrence Matrix (GLCM) features. The former reflects the microscopic texture of the road surface, while the latter reflects the contrast and correlation of the texture; the combination of the two can comprehensively characterize the visual characteristics of the road surface. Subsequently, a feature database containing common materials such as asphalt, cement, gravel, and soil is constructed. This database stores ultrasonic reflection feature samples and image texture feature samples corresponding to each material. The extracted features to be identified are matched with the database samples based on similarity. The material corresponding to the sample with the highest matching degree is selected as the current road surface material, and the light absorption rate parameter of that material is retrieved (e.g., gravel road surface absorption rate is approximately 40%, asphalt road surface approximately 20%). The system presets the expected road surface illumination intensity required for vehicle travel (effective illumination to ensure clear visibility for the driver). Based on the formula "target emitted light intensity = expected road surface illumination intensity / (1 - material light absorption rate)," the required light intensity output by the headlights is calculated, thus generating a second illumination adjustment command. For example, when the road surface is identified as gravel and the expected road surface illumination intensity is 800 lm, the target emitted light intensity needs to reach approximately 1333 lm. The command will specify this intensity parameter to ensure that different road surface materials ultimately achieve a consistent and effective lighting effect.

[0034] The data processing module is also used to fuse the vehicle speed signal and the vehicle body lateral acceleration data, and use a decision tree algorithm to determine whether the vehicle has entered a turning state. If it is determined to be turning, the steering detection unit and the tilt detection unit are triggered to switch from low power standby mode to high frequency working mode, receive and process steering angle data and tilt angle data, calculate the illumination angle adjustment value and illumination range expansion value in combination with preset fixed illumination area parameters, and generate a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value.

[0035] Specifically, in vehicle turning scenarios, the data processing module integrates vehicle speed signals and lateral acceleration data, along with steering and tilt angle data, to generate a third lighting adjustment command adapted to the turning posture. The core objective is to ensure that the headlights always illuminate a fixed area in front during a turn. Determining the turning state is the first step: the system first acquires vehicle speed signals from the vehicle's CAN bus and lateral acceleration data from the tilt detection unit, performing time synchronization and noise filtering to eliminate interference from bumps and sensor vibrations. Then, it constructs a decision tree system that includes vehicle speed characteristics, lateral motion characteristics, and the duration of the characteristic state. Through progressive logic, it first determines whether the vehicle speed is within a turning range (e.g., 0-25 km / h), then checks whether the lateral acceleration reaches the centrifugal force threshold corresponding to a turn, and finally verifies whether the state lasts for a preset duration (e.g., 0.5s). If all three conditions are met, it is determined to be a "turning state." After determining a turn, the system triggers the steering and tilt detection units to switch to high-frequency operating modes, collecting the handlebar steering angle θ and the vehicle tilt angle γ. During the parameter calculation phase, based on the principles of geometric optics, the horizontal steering angle adjustment value α is set to θ / 2. For example, when the handlebars turn 30° to the right, the headlights simultaneously turn 15° to the right, ensuring that the lighting direction matches the turning path. The vertical tilt angle adjustment value δ is calculated using the formula δ=arctan(h / (L0cosγ)) (where h is the headlight mounting height and L0 is the lighting length when driving straight), compensating for the vertical offset of the lighting area caused by the vehicle's tilt. The illumination range extension value β is equal to the steering angle θ, extending the corresponding angle towards the turning side to cover the lateral blind spot when turning. Finally, α, δ, β, and their corresponding adjustment directions and magnitudes are written into the third illumination adjustment command, clarifying the headlight angle and range adjustment parameters to ensure that the illumination area during turning remains consistent with that during straight driving.

[0036] The control execution module is used to drive the vehicle headlight module to adjust the illumination angle, illumination range and illumination intensity according to each illumination adjustment command, so as to realize the vehicle headlight adjustment.

[0037] Specifically, the control execution module, acting as the system's "execution layer," is responsible for receiving three types of illumination adjustment commands generated by the data processing module. Through a logic of "unit-based execution + multi-command coordination," it drives the headlight module to precisely adjust the illumination angle, range, and intensity. This module includes an angle range adjustment unit and an intensity adjustment unit: the angle range adjustment unit adjusts the illumination angle and range by driving a stepper motor and a light shield motor; the intensity adjustment unit uses PWM pulse width modulation technology to control the drive current of the headlight source, thereby changing the illumination intensity. In a single-instruction execution scenario, upon receiving the first instruction, the angle range adjustment unit drives the light shield to move laterally to expand the illumination range, while the intensity adjustment unit smoothly increases the drive current to enhance the light intensity. Upon receiving the second instruction, the intensity adjustment unit calculates the difference between the current intensity and the target value based on the target emitted light intensity in the instruction, and gradually adjusts the current through closed-loop control to ensure that the actual intensity remains stable at the target value. Upon receiving the third instruction, the angle range adjustment unit first adjusts the horizontal and vertical angles of the headlights, and then drives the light shield to expand the range towards the turning side. If the angle and range adjustment causes a decrease in light intensity per unit area, the intensity adjustment unit simultaneously increases the current to compensate. When multiple instructions are triggered simultaneously (such as when the heart rate exceeds the limit while turning), the system adopts a "safety first" strategy, prioritizing the execution of the third instruction—because the angle and range adjustment during turning directly affects the field of vision coverage, which is a primary requirement for safe driving. After the angle and range are adjusted to the appropriate level, the intensity adjustment requirements in the first and second instructions are then superimposed, achieving coordinated optimization of lighting in multiple scenarios.

[0038] In summary, the entire system constructs a complete closed-loop regulation logic through the comprehensive perception of the sensor module, the precise decision-making of the data processing module, and the efficient execution of the control execution module. It deeply integrates the driver's physiological state, road environment, vehicle posture, and headlight illumination to achieve dynamic adaptation of lighting parameters under different driving scenarios, ultimately significantly improving the safety and comfort of electric vehicles driving at night.

[0039] In an optional implementation, see Figure 2 As shown, Figure 2 The diagram shows a second type of intelligent headlight adjustment system for electric vehicles provided in Embodiment 1 of the present invention. The sensor module includes a heartbeat detection unit 201, a road surface detection unit 202, a steering detection unit 203, and a tilt detection unit 204.

[0040] The heartbeat detection unit is installed in the area where the handlebars contact the palm, and is used to continuously collect the driver's heart rate data at a preset sampling frequency, and transmit the heart rate data to the data processing module in real time.

[0041] Specifically, the heart rate detection unit is mainly responsible for collecting the driver's physiological state data. It is installed in the area where the handlebars contact the palm and acquires photoplethysmography (PPG) signals through optical sensors. Although this type of signal contains heart rate information, it is highly susceptible to external interference. For example, driver hand tremors can cause pulse-like spikes in the signal, poor contact between the sensor and the skin can cause signal disconnection, and changes in ambient light can cause baseline drift. To solve these problems, the preprocessing process first removes non-physiological interference through a combination of "bandpass filtering + sliding window filtering": the bandpass filter sets a passband of 0.5~8Hz to accurately filter out high-frequency noise from ambient light above 8Hz and baseline drift caused by slow hand movements below 0.5Hz; the sliding window filter uses 5 sampling points (approximately 5 seconds of data) as a window, averages the data within the window, and smooths the spikes caused by momentary hand tremors into a continuous signal. For example, the abnormal sequence "75→92→76" caused by hand tremors can be corrected into a stable sequence "75→76→77" after processing. Next, signal normalization is performed. A peak detection algorithm identifies the peak and trough values ​​of the pulse wave in the PPG signal, corresponding to changes in blood flow during cardiac contraction and relaxation. Simultaneously, "disconnected signals" (signals whose amplitude suddenly drops to 0 and persists for more than two sampling points) are removed, ensuring only continuous and complete pulse wave cycles are retained. Finally, preliminary feature extraction is completed, and the instantaneous heart rate is calculated based on the interval between adjacent peaks (e.g., a 0.8-second interval corresponds to 75 beats per minute). The data is then formatted into a structured form of "timestamp + heart rate value" to avoid the excessive data volume and complex parsing issues associated with directly transmitting the raw PPG signal before being sent to the data processing module in real time.

[0042] The road surface detection unit includes an ultrasonic sensor and a camera sensor. The ultrasonic sensor is used to collect ultrasonic reflection signals from the road surface at a preset sampling frequency. The camera sensor is used to collect road surface image data at a preset frame rate. The ultrasonic sensor and the camera sensor achieve timestamp alignment of data acquisition through a synchronous trigger signal, and transmit the ultrasonic reflection signals and the road surface image data to the data processing module.

[0043] Specifically, the road surface detection unit uses a combination of ultrasonic sensors and camera sensors to collect road surface data from two dimensions: "physical reflection" and "visual texture." The preprocessing of both sensors must consider their respective data characteristics while ensuring spatiotemporal alignment. For the ultrasonic sensors, the raw signals collected are millivolt-level analog voltage signals, easily affected by air humidity and small stones. During preprocessing, a programmable gain amplifier (PGA) is used to dynamically amplify the weak signal by 20 to 100 times—increasing the gain when the signal attenuates significantly in rainy weather and decreasing it in dry environments to ensure the signal amplitude meets the requirements of subsequent processing. Then, a bandpass filter with a center frequency of 40kHz is used to filter out low-frequency noise generated by motor operation and high-frequency interference from wind noise, retaining only the effective signals related to road surface reflection. Subsequently, time-domain analysis is performed on the filtered signal to extract three core parameters: amplitude ratio, propagation time deviation, and pulse width variation coefficient. The amplitude ratio reflects the road surface's reflectivity, the propagation time deviation reflects the road surface's smoothness (positive deviation at potholes), and the pulse width variation coefficient corresponds to the road surface's roughness. Finally, these parameters are bound to timestamps to form structured data. The raw road surface images captured by camera sensors are often blurred due to nighttime glare, rain reflections, and shadows. Preprocessing first employs a combination of Gaussian filtering and median filtering for noise reduction: a 3×3 convolutional kernel Gaussian filter smooths the fine granular Gaussian noise generated by the sensor circuitry, while a 5×5 window median filter eliminates salt-and-pepper noise such as bright spots from oncoming headlights and stone shadows, avoiding the problem of a single filter being unable to handle both types of noise. Then, a histogram equalization algorithm stretches the image grayscale values ​​from a concentrated low-brightness range to the full range of 0-255, enhancing the contrast between road surface texture and background, for example, making the fine cracks in asphalt pavement and the graininess of gravel pavement clearer. To reduce interference from invalid data, a region of interest (ROI) of 1-3 meters directly in front of the vehicle is cropped, removing irrelevant areas such as the sky and front-end components of the vehicle, reducing subsequent computational load. In addition, the two sensors will achieve data alignment through synchronous trigger signals, marking the ultrasonic parameters and ROI images at the same time stamp as a group, ensuring that the subsequent data processing module can accurately associate the road surface reflection characteristics and texture features at the same time, thereby improving the accuracy of material recognition.

[0044] The steering detection unit is installed at the rotation axis of the handlebars and is a Hall effect angle sensor with a preset angle measurement range and measurement accuracy.

[0045] Specifically, the steering detection unit is installed at the handlebar pivot and uses a Hall effect angle sensor to collect steering angle data. However, the raw data is easily affected by handlebar vibration and sensor installation deviation, causing fluctuations in the angle value. The first step in preprocessing is zero-point calibration: when the system is powered on and initialized, the initial angle of the handlebars when the vehicle is traveling straight is automatically recorded and set to 0°. All subsequent angle data are corrected based on this. For example, if there is a 5° deviation in the sensor installation, the 5° angle originally collected when traveling straight will be corrected to 0° to avoid steering judgment errors caused by installation errors. Next, a Kalman filter algorithm is used to smooth the angle fluctuations. This algorithm combines the angle measurement value of the sensor with the auxiliary data of the integrated gyroscope to estimate the optimal true angle—when the handlebar vibration causes the original angle to fluctuate between 28° and 32°, it can be smoothed to a stable 30°, while preserving the dynamic change trend of the steering angle. For example, when the handlebars turn from 0° to 30°, the angle value increases continuously without jumps. Finally, the calibrated angle data is converted into a structured format of "angle value + steering direction", clearly indicating left / right steering and the specific angle to avoid direction confusion during data processing module parsing.

[0046] The tilt detection unit is installed near the vehicle's center of gravity. It is a MEMS inertial measurement unit that integrates a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer. It is used to measure the vehicle's tilt angle and attitude, and has a preset attitude measurement accuracy.

[0047] Specifically, the tilt detection unit is installed near the vehicle's center of gravity. It collects three types of raw data—accelerometer, angular velocity, and magnetic force—via a MEMS inertial measurement unit. Accurate vehicle tilt angles require multi-data fusion. During preprocessing, a complementary filtering algorithm is first used for data fusion: the static tilt angle from the three-axis accelerometer is suitable for steady-state measurement but has a slow dynamic response, while the dynamic angular velocity from the three-axis gyroscope is suitable for dynamic measurement but prone to drift. Combining these two methods, the accelerometer data is prioritized for slow tilts, while the gyroscope data is prioritized for rapid tilts, balancing static accuracy and dynamic response speed to obtain a stable tilt angle. Because of electromagnetic interference from motors and magnetic fields from metal components in the vehicle environment, magnetometer data can deviate. Therefore, an ellipse fitting calibration algorithm is used to correct the magnetometer readings, ensuring the accuracy of its auxiliary calibration attitude and preventing drift over long-term measurements (e.g., a 15° tilt angle gradually becoming 18° due to magnetic field interference). Meanwhile, a reasonable tilt range of -30° to 30° is set (corresponding to the normal turning tilt range of electric vehicles). When the collected angle exceeds this range (such as an instantaneous 40° caused by vehicle bumps), it will be judged as an outlier and removed to prevent extreme interference from affecting the judgment of the data processing module.

[0048] The steering detection unit and the tilt detection unit are in a low-power standby mode when the vehicle is not turning. When the vehicle enters a turning state, they are triggered to start, and collect the handlebar steering angle data and the vehicle body tilt angle data respectively and transmit them to the data processing module.

[0049] Specifically, all preprocessed data from each unit is ultimately transmitted to the data processing module in a structured format (such as JSON or binary frames). This format includes key information such as timestamps, sensor types, data types, and specific values ​​or features. For example, the heartbeat detection unit transmits "timestamp + heart rate value," the road surface detection unit transmits "timestamp + ultrasonic three parameters + ROI image," and the steering and tilt units transmit "timestamp + steering angle / tilt angle." This structured output eliminates the need for format conversion or invalid data filtering in the data processing module, allowing it to directly access and analyze the data. This significantly improves the efficiency and accuracy of subsequent instruction generation, laying a data foundation for the stable operation of the entire intelligent vehicle lighting adjustment system. In an optional implementation, determining heart rate variability characteristics based on the heart rate data and generating a first illumination adjustment command for adjusting light intensity and illumination range based on the heart rate variability characteristics includes: The heart rate data is filtered in real time to remove noise interference and obtain a heart rate numerical sequence.

[0050] Specifically, real-time filtering of heart rate data is the first step in the entire process. Its purpose is to construct a stable heart rate numerical sequence, providing a reliable data foundation for subsequent analysis. During heart rate data acquisition, it is easily affected by various interference factors. For example, driver hand tremors can cause instantaneous data jumps, the contact gap between the sensor and the skin can cause reading fluctuations, and electromagnetic interference generated by the vehicle's motor can cause baseline drift. If these noises are not processed, they will seriously affect the accuracy of subsequent feature extraction. Therefore, the system adopts a combined approach of "sliding window filtering + Kalman filtering": The sliding window filtering sets five sampling points as a window (corresponding to five seconds of data). By averaging the data within the window, it smooths out short-term impulse noise such as sudden hand tremors. For example, the abnormal sequence "72→90→73" caused by hand tremors can be corrected to a continuous "72→73→74" after filtering. The Kalman filter utilizes the gradual nature of heart rate changes (under normal physiological conditions, heart rate does not rise or fall sharply in a short period). Through an iterative prediction-update process, it estimates the optimal true heart rate value, effectively eliminating long-term baseline drift caused by factors such as sensor temperature changes. For example, if the sensor reading remains 2 beats / minute higher after heating, the Kalman filter can correct this deviation in real time. After dual filtering, the final output heart rate sequence (e.g., [72,73,75,88,95,102] beats / minute) is continuous and stable, laying a solid foundation for extracting heart rate change features.

[0051] Heart rate variation characteristics are calculated based on the heart rate numerical sequence, wherein the heart rate variation characteristics include instantaneous peak heart rate, duration of heart rate exceeding threshold, and heart rate rise rate.

[0052] Specifically, heart rate variability calculation is a crucial step connecting physiological data with regulatory needs. It quantifies the driver's physiological stress level through three types of characteristic parameters, comprehensively reflecting their response to the current driving environment. The calculation of the instantaneous peak heart rate is based on the maximum value in the real-time tracking and filtering heart rate sequence. This peak directly corresponds to the driver's immediate physiological response to a sudden situation. For example, if an obstacle suddenly appears ahead, the driver's heart rate may surge from 80 beats per minute to 102 beats per minute within one second. This instantaneous peak of 102 beats per minute directly reflects the stress impact caused by the sudden change in environment. Calculating the duration of sustained heart rate exceeding the threshold requires first pre-setting a normal heart rate threshold range for the driver (usually 60-100 beats / minute). Timing begins when the heart rate first exceeds the upper limit of the range (100 beats / minute), and the duration of sustained exceedance is accumulated. For example, if 95 beats / minute (close to the upper limit) and 102 beats / minute (exceeding the upper limit) occur consecutively for 2 seconds in the sequence, the duration of sustained exceedance is recorded as 2 seconds. This duration parameter reflects the persistence of the complex environmental state—the longer the duration, the longer the driver is in a stress state, and the more urgent the need for stronger lighting. Calculating the rate of heart rate rise requires selecting the segment of heart rate change from the "upper limit of the normal range" to the "current value," and calculating using the formula "(current heart rate - upper limit of normal range) / change time." For example, if it takes 1 second for the heart rate to rise from 100 beats / minute to 102 beats / minute, the rate of rise is 2 beats / minute. The rate of increase (in seconds) reflects the urgency of environmental changes. The faster the rate increases, the higher the likelihood of sudden environmental changes, and the faster the driver needs to respond to risks by enhancing lighting.

[0053] Preset heart rate threshold range and corresponding light adjustment rules: When the instantaneous peak heart rate exceeds the upper limit of the heart rate threshold range, or the heart rate continues to exceed the threshold for a preset duration, or the heart rate rise rate exceeds the preset rate, the heart rate change characteristics are determined to meet the adjustment conditions.

[0054] Specifically, in the adjustment condition judgment stage, by pre-setting threshold ranges and logical rules, it is clear "when an adjustment command needs to be generated," ensuring timely response to stress demands while avoiding false triggers. First, the preset threshold range needs to be comprehensively set based on human physiological characteristics and driving scenario requirements: the normal heart rate threshold range is set at 60-100 beats / minute, with 100 beats / minute as the stress trigger point because the heart rate of healthy adults in a resting state is mostly within this range; exceeding it indicates a possible stress response. The duration of sustained exceedance of the threshold is set at 2 seconds to exclude the influence of short-term fluctuations, such as when the driver accidentally raises their hand, causing a sensor malfunction and the heart rate momentarily exceeding the upper limit but falling back within 1 second; in this case, no adjustment is needed. The heart rate rise rate is set at 5 beats / minute. If the heart rate exceeds a certain threshold within a few seconds, it is considered a rapid stress response. This is because the heart rate rises slowly during normal activity, and only sudden environmental changes can cause a rapid increase in heart rate. The logical judgment rule uses an "OR logic" trigger adjustment. This means that if any one of the following conditions is met: the instantaneous peak heart rate exceeds the upper limit, the duration of exceeding the threshold reaches the target, or the rate of increase exceeds the preset rate, the current heart rate change is considered to meet the adjustment conditions. This design can cover different types of stress scenarios—whether it's an instantaneous peak exceeding the limit due to a sudden situation or a prolonged exceeding the limit due to a complex environment, it can trigger lighting adjustment in a timely manner. Furthermore, to further avoid false triggers, a fluctuation tolerance range of ±5 times / minute is set. For brief exceeding of the threshold (falling back within 1 second) or minor exceeding of the threshold (less than 5 times / minute above the upper limit), adjustment will not be triggered, avoiding frequent headlight adjustments due to minor physiological fluctuations, which would affect the driver's visual experience.

[0055] The first illumination adjustment command is generated based on the satisfied adjustment conditions: the greater the heart rate exceeds the threshold, the longer the duration of exceeding the threshold, or the faster the heart rate rises, the higher the increase ratio of light intensity and the greater the expansion angle of the illumination range in the first illumination adjustment command.

[0056] Specifically, in the first stage of generating illumination adjustment instructions, the core is to achieve a quantitative mapping between heart rate variability characteristics and illumination parameters, generating differentiated instructions based on the stress level while ensuring dynamic adaptability. Regarding the quantitative mapping rules, the system establishes a correspondence between heart rate variability characteristics and illumination parameters: for every 10% increase in the heart rate exceeding the threshold (e.g., from 10% to 20% above the upper limit), the light intensity increase ratio increases by 15%-20%, and the light range expansion angle increases by 8°-10%; for example, when the threshold exceeds by 20%, the intensity increases by 40% and the range expands by 20°. For every second the duration of the heart rate exceeding the threshold increases, the intensity increase ratio increases by 5%-8%, and the range expansion angle increases by 3°-5%; for example, when it lasts for 3 seconds, the intensity increases by an additional 15% and the range expands by an additional 10°. For every 2 beats per minute increase in the heart rate rise rate... Seconds, the intensity increase ratio increases by 10%-12%, and the range expansion angle increases by 6°-8%, for example, the rate is 6 times / minute. Within one second, the intensity is increased by an additional 30% and the range is expanded by an additional 18°. This multi-dimensional quantitative mapping ensures that the magnitude of the lighting adjustment is precisely matched with the stress level. In terms of the instruction content, the first lighting adjustment instruction must include clear execution parameters: "Target value of lighting intensity" specifies the final brightness that the headlights need to achieve, such as a target value of 1400lm after a 40% increase from the current intensity of 1000lm; "Target value of lighting range" specifies the final range of the headlight illumination angle, such as a target value of 85° after a 25° increase from the current range of 60°; "Adjustment rate" is set to complete the adjustment within 1 second to avoid flickering caused by sudden increases or decreases in brightness and to protect the driver's vision. In addition, the commands are dynamically adaptable. If the heart rate change characteristics continue to upgrade (such as the over-threshold range increasing from 10% to 30%), the data processing module will update the command parameters in real time and gradually increase the lighting intensity and range. For example, if the initial command intensity is increased by 20%, and the over-threshold range is subsequently detected to increase, the intensity increase ratio will be immediately updated to 40%, ensuring that the "physiological state - lighting parameters" are always synchronized and adapted, providing drivers with continuous lighting support that meets their needs.

[0057] In an optional implementation, the step of determining road surface texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data respectively, identifying the road surface material based on the road surface texture features and ultrasonic reflection features, and generating a second illumination adjustment command for adjusting the illumination intensity based on the road surface material and the light absorption rate of different materials includes: The ultrasonic reflected signal is preprocessed by amplifying the weak signal and filtering out noise with a bandpass filter to extract ultrasonic reflection features, which include the reflected signal amplitude ratio, propagation time deviation, and pulse width variation coefficient.

[0058] Specifically, ultrasonic reflection signals are crucial data reflecting the physical characteristics of the road surface. Before feature extraction, they require targeted preprocessing to eliminate interference. Because ultrasonic waves attenuate due to increased distance and environmental factors (such as rain), the original reflected signals are often only in the millivolt range. Therefore, a programmable gain amplifier (PGA) circuit is needed for dynamic amplification of 20-100 times, with the gain automatically adjusted according to the environment. For example, the gain is increased when signal attenuation is significant in rainy weather to ensure that weak signals can be effectively identified. Simultaneously, the motor operation during vehicle movement generates low-frequency noise (<30kHz), and other sound waves in the environment introduce high-frequency interference (>50kHz). These factors mask the effective signals reflected from the road surface. Therefore, a bandpass filter circuit with a center frequency of 40kHz is needed to precisely filter out irrelevant noise, retaining only the signal components directly related to the road surface reflection.

[0059] After preprocessing, the ultrasonic reflection signal needs further extraction of core features to quantify road surface characteristics. First, the amplitude ratio of the reflected signal is calculated, which is the ratio of the maximum amplitude of the reflected signal to the amplitude of the transmitted signal. This ratio, ranging from 0 to 1, directly reflects the road surface's reflectivity. For example, asphalt pavements have a dense structure and strong reflectivity, with a ratio typically between 0.6 and 0.7, while gravel pavements have a loose structure and weak reflectivity, with a ratio often between 0.4 and 0.5. Second, the propagation time deviation is calculated. Based on the sensor's installation height and detection angle, the theoretical propagation time of the ultrasonic wave is first determined, and then subtracted from the actual measured propagation time (the time difference from transmission to reception). If there are potholes in the road surface, the actual propagation time will be longer, resulting in a positive deviation; if there are bumps in the road surface, the actual propagation time will be shorter, resulting in a negative deviation. This parameter can be used to determine the road surface smoothness. Finally, the pulse width variation coefficient is calculated, which is the ratio of the standard deviation to the average value of the reflected signal pulse width. Rough road surfaces cause the ultrasonic wave reflection direction to be dispersed, resulting in large pulse width fluctuations. The coefficient is usually greater than 0.3, while the reflection direction of smooth road surfaces is relatively concentrated, the pulse width is stable, and the coefficient is generally less than 0.2. This is used to distinguish the roughness of the road surface.

[0060] The road surface image data is preprocessed by using Gaussian filtering to reduce noise and histogram equalization to enhance contrast. A convolutional neural network algorithm is then used to extract the texture features of the road surface image. These texture features include local binary mode (LBP) histogram features and contrast and correlation features of the gray-level co-occurrence matrix (GLCM).

[0061] Specifically, road surface image data supplements road surface information from the visual texture dimension, and also requires preprocessing to enhance feature recognition. During image acquisition, sensor circuits generate fine-grained Gaussian noise, while headlight glare and small pebble shadows create isolated bright spots or dark spots—salt-and-pepper noise. This noise interferes with texture feature extraction. Therefore, a 3×3 convolutional kernel Gaussian filtering algorithm is first used to smooth the Gaussian noise by weighted averaging of the area surrounding the pixel. Then, a 5×5 window median filtering algorithm is used to replace the center pixel with the median value of the pixels within the window, effectively eliminating salt-and-pepper noise. In addition, during nighttime driving, image grayscale values ​​are often concentrated in the low-brightness range, resulting in low contrast between road surface texture and background. A histogram equalization algorithm is needed to stretch the grayscale values ​​from the concentrated range to the full range of 0-255, making the texture details such as cracks and particles on the road surface clearer and preventing texture features from being masked due to insufficient contrast.

[0062] After preprocessing, the road surface image needs to have its texture features extracted using algorithms to quantify the micro and macro texture attributes of the road surface. Local Binary Pattern (LBP) histogram feature extraction involves comparing the grayscale value of each pixel with its eight neighboring pixels. If the grayscale value of a neighboring pixel is greater than that of the center pixel, it is recorded as 1; otherwise, it is recorded as 0, generating an 8-bit binary code. The distribution of all codes across the entire image is then statistically analyzed to form a histogram. This histogram reflects the roughness of the road surface's micro-texture. For example, the micro-particle distribution of gravel roads is uneven, resulting in a more dispersed histogram peak, while the micro-texture of asphalt roads is uniform, leading to a relatively concentrated histogram peak. The extraction of Gray-Level Co-occurrence Matrix (GLCM) features involves first setting the distance *d* and angle (0°, 45°, 90°, 135°) between pixel pairs, then calculating the joint distribution matrix of gray values ​​for pixel pairs satisfying these distance and angle conditions. Two key parameters, contrast and correlation, are then extracted from this matrix. Contrast reflects the clarity of the texture; a higher value indicates a more distinct texture outline. Correlation reflects the consistency of the texture direction; a higher value indicates a more uniform texture direction. For example, the texture of cement roads tends to have a regular direction and high correlation, while the texture of dirt roads is chaotic and has low correlation. These two parameters quantify the macroscopic texture features of the road surface. To balance feature extraction accuracy and computational efficiency, a lightweight MobileNet convolutional neural network is used to process the preprocessed image end-to-end, directly outputting a high-dimensional vector (dimensions between 64 and 256) containing LBP histogram features and GLCM features. This ensures feature integrity while avoiding the computational burden of complex algorithms.

[0063] A road surface material feature database is constructed, which stores ultrasonic reflection feature samples and road surface image texture feature samples corresponding to different road surface materials.

[0064] Specifically, the database contains road surface material characteristics. It includes samples of five typical road surface materials: asphalt, cement, gravel, soil, and snow / ice. Each material category has at least 1000 samples to ensure representativeness. Each sample is associated with three types of information: first, an ultrasonic reflection characteristic triplet, namely, the reflected signal amplitude ratio, propagation time deviation, and pulse width variation coefficient; second, a road surface image texture feature vector, including LBP histogram features and GLCM contrast and correlation features; and third, the light absorption rate parameter corresponding to the material obtained through experimental measurements, such as approximately 40% for gravel roads, approximately 20% for asphalt roads, and as high as 60% for snow / ice roads.

[0065] The extracted ultrasonic reflection features and road surface image texture features are matched with samples in the road surface material feature database. The material corresponding to the sample with the highest matching degree is selected as the identified road surface material, and the light absorption rate parameter corresponding to the road surface material is retrieved.

[0066] Specifically, the extracted features to be identified first need to be normalized. Both the numerical features of ultrasound and the vector features of the image are standardized to the [0,1] interval to eliminate the influence of differences in the dimensions of different features. For example, the amplitude ratio of reflected signals ranges from 0 to 1, while the propagation time deviation may be on the order of milliseconds. Normalization allows both types of features to have equal weight in subsequent calculations. Then, a weighted Euclidean distance algorithm is used to calculate the similarity between the features to be identified and the features of each sample in the database. Considering that visual texture contributes more to material identification, ultrasound features are assigned a weight of 0.4, and image features are assigned a weight of 0.6. This weighted calculation reduces the distance between the features to be identified and the sample features; the smaller the distance, the higher the similarity. Finally, material determination is performed. The material corresponding to the sample with the highest similarity (over 85%) is selected as the final identification result. If the highest similarity is below 60%, it indicates that the current road surface material is not in the preset typical category and is treated as "unknown material" by default. Conservative light intensity parameters are used to avoid improper lighting adjustment due to inaccurate material identification.

[0067] The expected road surface illumination intensity required for vehicle operation is preset. Based on the light absorption rate parameter corresponding to the road surface material and the expected road surface illumination intensity, the target emitted light intensity to be output by the headlight module is determined, and the second illumination adjustment command is generated.

[0068] Specifically, the generation of the second illumination adjustment command is based on calculating the target emission intensity of the headlights according to the identified road surface material and its light absorption rate. First, based on the visual needs of the human eye at different vehicle speeds, the effective road illumination intensity is preset. For example, when the vehicle is traveling at low speed (20km / h), the driver has a higher need to observe the details of the road surface at close range, so the expected road illumination intensity is set to 800lm. When traveling at high speed (50km / h), a greater field of vision is required, so the expected road illumination intensity is increased to 1200lm to ensure that the driver can clearly identify road obstacles.

[0069] According to optical principles, the actual effective light intensity received by the road surface is equal to the light intensity emitted by the headlight multiplied by (1 - light absorption rate of the road surface material). Therefore, the required light intensity of the headlight needs to be calculated by using the formula "target emitted light intensity = expected road surface light intensity / (1 - light absorption rate)". For example, when the road surface is identified as gravel material (light absorption rate 40%) and the expected road surface light intensity is 800lm, the target emitted light intensity = 800 / (1 - 40%) ≈ 1333lm.

[0070] The final generated second illumination adjustment command must include specific execution parameters: first, a target emission intensity value, such as 1333 lm, providing a target benchmark for intensity adjustment to the control execution module; second, an adjustment step size, typically set to adjust by 50 lm every 100 ms to avoid visual discomfort to the driver due to sudden changes in illumination intensity; and third, a duration, i.e., the length of time to maintain the target emission intensity until the system updates the road surface material recognition results, ensuring that the illumination intensity can be adjusted promptly according to changes in road surface material. After receiving the command, the control execution module adjusts the LED drive current using PWM (Pulse Width Modulation) technology. For example, a target emission intensity of 1333 lm corresponds to a drive current of 1.3A, thereby precisely controlling the illumination intensity of the headlights to ensure that different road surface materials ultimately receive consistent and effective illumination, improving nighttime driving safety.

[0071] In an optional implementation, the fusion of vehicle speed signal and lateral acceleration data, and the use of a decision tree algorithm to determine whether the vehicle has entered a turning state, includes: The system acquires vehicle speed signals and lateral acceleration data during vehicle operation, and performs time synchronization and noise filtering on the two types of data.

[0072] Specifically, determining a vehicle's turning state begins with acquiring and preprocessing vehicle speed signals and lateral acceleration data. First, vehicle speed signals are acquired from the vehicle's CAN bus or wheel speed sensors, while lateral acceleration data is extracted from the triaxial accelerometer of the tilt detection unit; both are sampled at a frequency of 10Hz. To ensure accuracy, these two types of data undergo time synchronization. Timestamps are added to the data based on the system's unified clock, and linear interpolation is used to align any potential timing discrepancies, ensuring a one-to-one correspondence between vehicle speed and lateral acceleration at the same moment. Noise filtering is then applied. A first-order low-pass filter (cutoff frequency 1Hz) is used on the vehicle speed signal to smooth out instantaneous jumps caused by bumps, while a sliding window mean filter (window size of 3 sampling points) is used on the lateral acceleration data to eliminate high-frequency vibration noise, preserving the stable centrifugal acceleration component during turning, thus providing a reliable data foundation for subsequent judgments.

[0073] A decision tree algorithm is constructed to establish a feature system, which includes key features related to vehicle turning behavior and corresponding decision thresholds. The key features include at least vehicle speed features, vehicle lateral motion features, and feature state duration.

[0074] Specifically, the next step is to construct a judgment feature system for the decision tree algorithm. This system includes key features closely related to vehicle turning behavior and corresponding judgment thresholds. Among these, the vehicle speed feature uses 0-25 km / h as the easy-to-turn speed range, which is determined based on the dynamic characteristics of two-wheeled electric vehicles. At low speeds, steering agility is high, while at high speeds, the vehicle mostly maintains a straight-line state. For the lateral motion feature, a lateral acceleration ≥0.6g is set as the turning judgment threshold, corresponding to the characteristic lateral acceleration generated by centrifugal force during turning. The duration of the feature state is ≥0.5s to distinguish between genuine turning and feature changes caused by instantaneous operations (such as brief turns to avoid obstacles), thus avoiding misjudgments.

[0075] Based on the aforementioned judgment feature system, a progressive decision tree judgment logic is established: first, the vehicle speed feature is used to determine whether the vehicle is in a speed range that is easy to turn; then, the vehicle body lateral movement feature is used to determine whether the vehicle body generates lateral movement related to turning; finally, the duration of the feature state is used to determine whether the above features constitute a stable turning-related state.

[0076] Specifically, based on the aforementioned feature system, a progressive decision tree judgment logic is established. First, it determines whether the vehicle's speed is within the easy-to-turn range of 0-25 km / h. If the speed exceeds this range, it is directly judged as a non-turning state. If the speed is within this range, it further checks whether the vehicle's lateral acceleration is ≥0.6g. If it does not reach this threshold, the non-turning judgment is maintained. Once the lateral acceleration meets the threshold, it finally verifies whether this state lasts ≥0.5s. Only when the speed range, lateral acceleration threshold, and duration requirements are simultaneously met is it judged as a turning state. This progressive logic, through layer-by-layer filtering, gradually focuses on the feature combinations of actual turning behavior, improving the accuracy of the judgment. The decision tree algorithm is used to comprehensively analyze the judgment results of the above features and output a judgment on whether the vehicle has entered a turning state.

[0077] Specifically, the decision tree algorithm comprehensively analyzes the judgment results of the above features and outputs a judgment on whether the vehicle has entered a turning state. Based on this judgment, the corresponding operation is executed.

[0078] Based on the judgment result, the corresponding operation is triggered: if it is determined that the vehicle has entered a turning state, a trigger signal is generated to control the steering detection unit and tilt detection unit to switch from low power standby mode to high frequency working mode; if it is determined that the vehicle is not in a turning state, the low power standby mode of the steering detection unit and tilt detection unit is maintained.

[0079] Specifically, if a turning state is determined, a trigger signal is immediately generated to control the steering detection unit and tilt detection unit to switch from low-power standby mode (current ≤5μA and ≤8μA respectively) to high-frequency operating mode (sampling frequency increased to 100Hz) to collect steering angle and tilt angle data in real time; if a non-turning state is determined, the two detection units are kept in low-power standby mode, and the system energy consumption is optimized while ensuring data collection needs and extending the electric vehicle's range by activating them on demand.

[0080] In an optional implementation, the step of calculating the illumination angle adjustment value and the illumination range expansion value based on preset fixed illumination area parameters, and generating a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value, includes: The preset fixed lighting area parameters include the ideal area length L0, width W0, and headlight installation height h of the area where the headlights illuminate the ground in front of the vehicle when it is traveling straight.

[0081] Specifically, the generation of the third illumination adjustment command begins with preset fixed illumination area parameters. These parameters are fundamental to ensuring stable lighting performance during cornering, including the ideal length L0 and width W0 of the area illuminated by the headlights when the vehicle is traveling straight, as well as the headlight mounting height h. These parameters are preset according to vehicle design specifications and safety lighting requirements; for example, L0 can be set to 5 meters, W0 to 2 meters, and h to 0.8 meters to ensure sufficient safe field of vision when traveling straight. The system receives and processes the handlebar steering angle data θ collected by the steering detection unit and the vehicle body tilt angle data γ collected by the tilt detection unit.

[0082] Specifically, after the vehicle is determined to be turning, the system receives handlebar steering angle data θ collected by the steering detection unit and body tilt angle data γ collected by the tilt detection unit. This data is processed in real time to ensure it accurately reflects the vehicle's current steering amplitude and tilt state. The processing includes removing high-frequency noise from the data and smoothing instantaneous fluctuations using filtering algorithms, enabling θ and γ to stably reflect the vehicle's steering and tilt trends, providing reliable parameters for subsequent calculations.

[0083] The illumination angle adjustment value is calculated based on the principle of geometric optics, including the horizontal turning angle adjustment value α=θ / 2 and the vertical tilt angle adjustment value δ=arctan(h / (L0cosγ)).

[0084] Specifically, based on the principles of geometric optics, the system calculates the illumination angle adjustment value. The horizontal steering angle adjustment value α is set to half the handlebar steering angle θ, i.e., α = θ / 2. This design ensures that the headlight illumination direction matches the vehicle's turning path, preventing the illumination area from deviating from the actual direction of travel due to excessive handlebar steering. For example, when the handlebars turn 30° to the right, the headlight simultaneously turns 15° to the right horizontally, ensuring illumination of the critical path on the inside of the turn. The vertical tilt angle adjustment value δ is calculated using the formula δ = arctan(h / (L0cosγ)). This formula considers the influence of the vehicle's tilt angle γ on vertical illumination. When the vehicle tilts, the change in cosγ value leads to adjustments in the calculation results, causing the headlight to tilt accordingly in the vertical direction. This compensates for the vertical shift of the illumination area caused by the vehicle's tilt, ensuring that the illumination area always falls at the ideal height above the ground.

[0085] The illumination range extension value β=θ is determined based on the handlebar steering angle θ, and the illumination range is extended towards the steering side.

[0086] Specifically, the illumination range extension value β is determined based on the handlebar steering angle θ, i.e., β=θ, and the extension direction is consistent with the steering direction. This means that the larger the handlebar steering angle, the larger the illumination range extends to the steering side. For example, when the handlebar turns 30° to the right, the illumination range on the right side expands by 30° accordingly to cover the blind spot in the side view that occurs when turning, ensuring that the driver can clearly observe the road conditions on the steering side.

[0087] The horizontal steering angle adjustment value α, the vertical tilt angle adjustment value δ, and the illumination range expansion value β are written into the third illumination adjustment command, and the direction, magnitude, and range expansion direction and angle of the headlight angle adjustment are specified.

[0088] Specifically, the system finally writes the calculated horizontal steering angle adjustment value α, vertical tilt angle adjustment value δ, and illumination range expansion value β into the third illumination adjustment command. It also specifies the exact direction of the headlight angle adjustment (e.g., left or right, up or down), the adjustment range, and the direction and angle of the illumination range expansion. This command allows the control execution module to precisely drive the headlight adjustment mechanism, achieving dynamic adaptation of the illumination angle and range. This ensures that the illumination area remains consistent with the fixed illumination area parameters during straight-line driving, providing the driver with a stable and clear field of vision.

[0089] In an optional implementation, the step of driving the headlight module to adjust the illumination angle, illumination range, and illumination intensity according to each illumination adjustment command to achieve headlight adjustment includes: The control execution module is equipped with an angle range adjustment unit and an intensity adjustment unit. The angle range adjustment unit is used to drive the headlight module to adjust the illumination angle and illumination range, and the intensity adjustment unit is used to adjust the illumination intensity of the headlight through current control.

[0090] Specifically, the control execution module achieves precise control of the headlight module through the angle range adjustment unit and the intensity adjustment unit. The angle range adjustment unit is responsible for driving the headlight to adjust the illumination angle and range, while the intensity adjustment unit adjusts the illumination intensity through current control. The two work together to respond to different illumination adjustment commands.

[0091] When the first illumination adjustment command is received, the angle range adjustment unit drives the headlight module to expand the illumination coverage range, and the intensity adjustment unit improves the illumination intensity of the headlight by smoothly adjusting the driving current of the headlight source.

[0092] Specifically, upon receiving the first illumination adjustment command, the angle range adjustment unit immediately activates, driving the stepper motor and the light shield motor to expand the illumination coverage of the headlights to both sides, with the expansion angle strictly following the set value in the command. Simultaneously, the intensity adjustment unit employs a smooth adjustment strategy, gradually changing the driving current of the LED light source (the adjustment rate is controlled to increase or decrease by no more than 200mA every 100ms) to achieve a steady increase in illumination intensity, avoiding light flickering caused by sudden current changes and ensuring the driver's visual comfort. When the second illumination adjustment command is received, the intensity adjustment unit calculates the difference between the current illumination intensity and the target illumination intensity based on the target emitted illumination intensity determined in the second illumination adjustment command, and gradually adjusts the illumination intensity to the target value through a closed-loop control strategy.

[0093] Specifically, for the second illumination adjustment command, the intensity adjustment unit first calculates the difference between the current illumination intensity and the target emitted illumination intensity. Then, it adopts a PID closed-loop control strategy to dynamically adjust the duty cycle of the PWM signal according to the magnitude of the difference (with an adjustment accuracy of 1%). By continuously feeding back the current actual illumination intensity (monitored in real time by the current sampling resistor), the illumination intensity is gradually corrected to the target value. The entire adjustment process is usually completed within 1-2 seconds, and the overshoot is controlled within 5%, ensuring the stability and accuracy of the adjustment.

[0094] When the third illumination adjustment command is received, the angle range adjustment unit drives the headlight module to adjust the horizontal and vertical illumination angles, and at the same time drives the headlight module to expand the illumination range towards the turning side; if the angle and range adjustment causes a decrease in the illumination intensity per unit area, the intensity adjustment unit synchronously compensates and adjusts the illumination intensity.

[0095] Specifically, upon receiving the third illumination adjustment command, the angle range adjustment unit first drives the horizontal stepper motor to change the horizontal illumination angle of the headlights according to the horizontal steering angle adjustment value α. Simultaneously, it adjusts the vertical angle of the headlights to the value δ via the vertical stepper motor. Simultaneously, it drives the light shield motor on the steering side to extend the illumination range towards the steering side by an angle β. Since angle and range adjustments may cause a decrease in illumination intensity per unit area (e.g., energy dispersion after the illumination range expands), the intensity adjustment unit calculates a compensation coefficient in real time (based on the proportion of change in the illuminated area) and performs synchronous compensation by increasing the drive current to ensure that the actual illumination intensity received by the road surface remains stable. When multiple illumination adjustment commands are triggered simultaneously, the control execution module responds first to the third illumination adjustment command to complete the adjustment of the illumination angle and range; then, based on the adjusted angle and range, it superimposes the illumination intensity adjustment requirements from other illumination adjustment commands.

[0096] Specifically, if multiple illumination adjustment commands are triggered simultaneously, the control execution module will prioritize responding to the third illumination adjustment command, prioritizing the adjustment of the illumination angle and range. This is because the angle adaptation during turning directly affects the driver's observation of the turning path, which is a primary requirement for safe driving. After the angle and range are adjusted to the correct position, the illumination intensity adjustment requirements from the first and second illumination adjustment commands are superimposed on the current illumination state. By calculating the superposition ratio of the intensity adjustments of each command (e.g., if the first command requires an increase of 40% and the second command requires an increase of 30%, then the final superposition increase is 70%), the coordinated execution of multiple commands is achieved, ensuring that the headlights can still provide optimal lighting effects in complex scenarios.

[0097] Example 2 See Figure 3 As shown, Figure 3 The flowchart illustrates a method for adjusting the intelligent headlights of an electric vehicle according to Embodiment 2 of the present invention. This method is applied to an intelligent headlight adjustment system for an electric vehicle, which includes a sensor module, a data processing module, a control execution module, and a headlight module. The method includes steps S301-S305: S301: The sensor module acquires the driver's heart rate data, ultrasonic reflection signals from the road surface, road image data, and vehicle handlebar steering angle data and vehicle body tilt angle data. S302: The data processing module determines the heart rate change characteristics based on the heart rate data, and generates a first light adjustment command for adjusting the light intensity and light range based on the heart rate change characteristics; S303: The data processing module determines the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, identifies the road surface material based on the road surface image texture features and ultrasonic reflection features, and generates a second light adjustment command for adjusting the light intensity based on the road surface material and the light absorption rate of different materials. S304: The data processing module integrates the vehicle speed signal and the vehicle body lateral acceleration data, and uses a decision tree algorithm to determine whether the vehicle has entered a turning state; if it is determined to be turning, it triggers the steering detection unit and the tilt detection unit to switch from low power standby mode to high frequency working mode, receives and processes steering angle data and tilt angle data, calculates the illumination angle adjustment value and illumination range expansion value in combination with preset fixed illumination area parameters, and generates a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value; S305: The control execution module drives the vehicle headlight module to adjust the illumination angle, illumination range and illumination intensity according to each illumination adjustment command, so as to realize the vehicle headlight adjustment.

[0098] In one optional implementation, the sensor module includes a heartbeat detection unit, a road surface detection unit, a steering detection unit, and a tilt detection unit; the heartbeat detection unit is installed in the area where the handlebars contact the palm; the road surface detection unit includes an ultrasonic sensor and a camera sensor; the steering detection unit is installed at the rotation axis of the handlebars and is a Hall effect angle sensor with a preset angle measurement range and measurement accuracy; the tilt detection unit is installed near the vehicle's center of gravity and is a MEMS inertial measurement unit integrating a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer, with a preset attitude measurement accuracy; The method includes: The heart rate detection unit continuously collects the driver's heart rate data at a preset sampling frequency and transmits the heart rate data to the data processing module in real time. The ultrasonic sensor collects ultrasonic reflection signals from the road surface at a preset sampling frequency. The camera sensor acquires road image data at a preset frame rate; The ultrasonic sensor and the camera sensor achieve data acquisition timestamp alignment through a synchronous trigger signal, and transmit the ultrasonic reflected signal and the road image data to the data processing module; The tilt detection unit measures the vehicle body tilt angle and attitude; The steering detection unit and the tilt detection unit are in a low-power standby mode when the vehicle is not turning. When the vehicle enters a turning state, they are triggered to start, and collect the handlebar steering angle data and the vehicle body tilt angle data respectively and transmit them to the data processing module.

[0099] In an optional implementation, determining heart rate variability characteristics based on the heart rate data and generating a first illumination adjustment command for adjusting light intensity and illumination range based on the heart rate variability characteristics includes: The heart rate data is filtered in real time to remove noise interference and obtain a heart rate numerical sequence; Heart rate change characteristics are calculated based on the heart rate numerical sequence, wherein the heart rate change characteristics include instantaneous peak heart rate, duration of heart rate exceeding threshold, and rate of heart rate rise. Preset heart rate threshold range and corresponding light adjustment rules: When the instantaneous peak heart rate exceeds the upper limit of the heart rate threshold range, or the heart rate continues to exceed the threshold for a preset duration, or the heart rate rise rate exceeds the preset rate, the heart rate change characteristics are determined to meet the adjustment conditions. The first illumination adjustment command is generated based on the satisfied adjustment conditions: the greater the heart rate exceeds the threshold, the longer the duration of exceeding the threshold, or the faster the heart rate rises, the higher the increase ratio of light intensity and the greater the expansion angle of the illumination range in the first illumination adjustment command.

[0100] In an optional implementation, the step of determining road surface texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data respectively, identifying the road surface material based on the road surface texture features and ultrasonic reflection features, and generating a second illumination adjustment command for adjusting the illumination intensity based on the road surface material and the light absorption rate of different materials includes: The ultrasonic reflected signal is preprocessed by amplifying the weak signal and filtering out noise with a bandpass filter to extract ultrasonic reflection features, which include the reflected signal amplitude ratio, propagation time deviation and pulse width variation coefficient. The road surface image data is preprocessed by Gaussian filtering algorithm for noise reduction and histogram equalization algorithm for contrast enhancement. Convolutional neural network algorithm is used to extract the texture features of the road surface image. The texture features of the road surface image include local binary mode (LBP) histogram features and contrast and correlation features of gray-level co-occurrence matrix (GLCM). Construct a road surface material feature database, which stores ultrasonic reflection feature samples and road surface image texture feature samples corresponding to different road surface materials; The extracted ultrasonic reflection features and road surface image texture features are matched with samples in the road surface material feature database. The material corresponding to the sample with the highest matching degree is selected as the identified road surface material, and the light absorption rate parameter corresponding to the road surface material is retrieved. The expected road surface illumination intensity required for vehicle operation is preset. Based on the light absorption rate parameter corresponding to the road surface material and the expected road surface illumination intensity, the target emitted light intensity to be output by the headlight module is determined, and the second illumination adjustment command is generated.

[0101] In an optional implementation, the fusion of vehicle speed signal and lateral acceleration data, and the use of a decision tree algorithm to determine whether the vehicle has entered a turning state, includes: The vehicle speed signal and lateral acceleration data during the vehicle's movement are acquired, and the two types of data are processed for time synchronization and noise filtering. Construct a judgment feature system for the decision tree algorithm. The feature system includes key features related to vehicle turning behavior and corresponding judgment thresholds. The key features include at least vehicle speed features, vehicle lateral motion features, and feature state duration. Based on the aforementioned judgment feature system, a progressive decision tree judgment logic is established: first, the vehicle speed feature is used to determine whether the vehicle is in a speed range that is easy to turn; then, the vehicle body lateral movement feature is used to determine whether the vehicle body generates lateral movement related to turning; finally, the duration of the feature state is used to determine whether the above features constitute a stable turning-related state. The decision tree algorithm is used to comprehensively analyze the judgment results of the above features and output a judgment on whether the vehicle has entered a turning state. Based on the judgment result, the corresponding operation is triggered: if it is determined that the vehicle has entered a turning state, a trigger signal is generated to control the steering detection unit and tilt detection unit to switch from low power standby mode to high frequency working mode; if it is determined that the vehicle is not in a turning state, the low power standby mode of the steering detection unit and tilt detection unit is maintained.

[0102] In an optional implementation, the step of calculating the illumination angle adjustment value and the illumination range expansion value based on preset fixed illumination area parameters, and generating a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value, includes: Preset fixed lighting area parameters, wherein the fixed lighting area parameters include the ideal area length L0, width W0, and headlight installation height h of the area illuminated by the headlights when the vehicle is traveling straight; Receive and process the handlebar steering angle data θ collected by the steering detection unit and the vehicle body tilt angle data γ collected by the tilt detection unit; The illumination angle adjustment value is calculated based on the principle of geometric optics, including the horizontal turning angle adjustment value α=θ / 2 and the vertical tilt angle adjustment value δ=arctan(h / (L0cosγ)); The illumination range extension value β=θ is determined based on the handlebar steering angle θ, and the illumination range is extended towards the steering side; The horizontal steering angle adjustment value α, the vertical tilt angle adjustment value δ, and the illumination range expansion value β are written into the third illumination adjustment command, and the direction, magnitude, and range expansion direction and angle of the headlight angle adjustment are specified.

[0103] In an optional implementation, the step of driving the headlight module to adjust the illumination angle, illumination range, and illumination intensity according to each illumination adjustment command to achieve headlight adjustment includes: The control execution module is configured with an angle range adjustment unit and an intensity adjustment unit. The angle range adjustment unit is used to drive the vehicle lamp module to adjust the illumination angle and illumination range, and the intensity adjustment unit is used to adjust the illumination intensity of the vehicle lamp through current control. When the first illumination adjustment command is received, the angle range adjustment unit drives the headlight module to expand the illumination coverage range, and the intensity adjustment unit improves the illumination intensity of the headlight by smoothly adjusting the driving current of the headlight source. When the second illumination adjustment command is received, the intensity adjustment unit calculates the difference between the current illumination intensity and the target illumination intensity based on the target emitted illumination intensity determined in the second illumination adjustment command, and gradually adjusts the illumination intensity to the target value through a closed-loop control strategy. When the third illumination adjustment command is received, the angle range adjustment unit drives the headlight module to adjust the horizontal and vertical illumination angles, and at the same time drives the headlight module to expand the illumination range towards the turning side; if the angle and range adjustment causes a decrease in the illumination intensity per unit area, the intensity adjustment unit synchronously compensates and adjusts the illumination intensity. When multiple illumination adjustment commands are triggered simultaneously, the control execution module responds first to the third illumination adjustment command to complete the adjustment of the illumination angle and range; then, based on the adjusted angle and range, it superimposes the illumination intensity adjustment requirements from other illumination adjustment commands.

[0104] Example 3 Based on the same application concept, see [link / reference] Figure 4 As shown, Figure 4 A schematic diagram of the structure of a computer device provided in Embodiment 3 of the present invention is shown, wherein, as Figure 4 As shown, the computer device 400 provided in Embodiment 3 of this application includes: The system includes a processor 401, a memory 402, and a bus 403. The memory 402 stores machine-readable instructions that can be executed by the processor 401. When the computer device 400 is running, the processor 401 communicates with the memory 402 through the bus 403. When the machine-readable instructions are executed by the processor 401, the steps of the electric vehicle intelligent headlight adjustment method shown in Embodiment 2 are performed.

[0105] Example 4 Based on the same concept, this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the electric vehicle intelligent headlight adjustment method described in any of the above embodiments.

[0106] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system and apparatus described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0107] The computer program product for intelligent headlight adjustment of electric vehicles provided in this embodiment of the invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods described in the preceding method embodiments. For specific implementation details, please refer to the method embodiments, which will not be repeated here.

[0108] The intelligent headlight adjustment system for electric vehicles provided in this embodiment of the invention can be specific hardware on the device or software or firmware installed on the device. The method provided in this embodiment of the invention has the same implementation principle and technical effects as the aforementioned method embodiments. For the sake of brevity, any parts not mentioned in the method embodiments can be referred to the corresponding content in the aforementioned system embodiments. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the methods described above can all be referred to the corresponding processes in the above system embodiments, and will not be repeated here.

[0109] In the embodiments provided by this invention, it should be understood that the disclosed systems and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the shown or discussed mutual couplings, direct couplings, or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0110] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0111] In addition, the functional units in the embodiments provided by the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0112] If the aforementioned functions are implemented as 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 this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0113] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0114] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. All should be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An intelligent headlight adjustment system for electric vehicles, characterized in that, The system includes a sensor module, a data processing module, a control execution module, and a vehicle lighting module; The sensor module is used to acquire the driver's heart rate data, ultrasonic reflection signals from the road surface, road image data, and vehicle handlebar steering angle data and vehicle body tilt angle data. The data processing module is used to determine the heart rate change characteristics based on the heart rate data, and generate a first light adjustment command for adjusting the light intensity and light range based on the heart rate change characteristics. The data processing module is further configured to determine the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, respectively, identify the road surface material based on the road surface image texture features and ultrasonic reflection features, and generate a second light adjustment command for adjusting the light intensity based on the road surface material and the light absorption rate of different materials. The data processing module is also used to fuse the vehicle speed signal and the lateral acceleration data of the vehicle body, and use a decision tree algorithm to determine whether the vehicle has entered a turning state; if it is determined to be turning, the steering detection unit and the tilt detection unit are triggered to switch from low power standby mode to high frequency working mode, receive and process steering angle data and tilt angle data, calculate the illumination angle adjustment value and illumination range expansion value in combination with preset fixed illumination area parameters, and generate a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value; The control execution module is used to drive the vehicle headlight module to adjust the illumination angle, illumination range and illumination intensity according to each illumination adjustment command, so as to realize the vehicle headlight adjustment.

2. The intelligent headlight adjustment system for electric vehicles according to claim 1, characterized in that, The sensor module includes a heartbeat detection unit, a road surface detection unit, a steering detection unit, and a tilt detection unit; The heartbeat detection unit is installed in the area where the handlebars contact the palm, and is used to continuously collect the driver's heart rate data at a preset sampling frequency, and transmit the heart rate data to the data processing module in real time. The road surface detection unit includes an ultrasonic sensor and a camera sensor. The ultrasonic sensor is used to collect ultrasonic reflection signals from the road surface at a preset sampling frequency. The camera sensor is used to collect road surface image data at a preset frame rate. The ultrasonic sensor and the camera sensor achieve timestamp alignment of data acquisition through a synchronous trigger signal, and transmit the ultrasonic reflection signals and the road surface image data to the data processing module. The steering detection unit is installed at the rotation axis of the handlebars and is a Hall effect angle sensor with a preset angle measurement range and measurement accuracy. The tilt detection unit is installed near the vehicle's center of gravity. It is a MEMS inertial measurement unit that integrates a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer. It is used to measure the vehicle's tilt angle and attitude, and has a preset attitude measurement accuracy. The steering detection unit and the tilt detection unit are in a low-power standby mode when the vehicle is not turning. When the vehicle enters a turning state, they are triggered to start, and collect the handlebar steering angle data and the vehicle body tilt angle data respectively and transmit them to the data processing module.

3. The intelligent headlight adjustment system for electric vehicles according to claim 1, characterized in that, The step of determining heart rate change characteristics based on the heart rate data and generating a first light adjustment command for adjusting light intensity and light range based on the heart rate change characteristics includes: The heart rate data is filtered in real time to remove noise interference and obtain a heart rate numerical sequence; Heart rate change characteristics are calculated based on the heart rate numerical sequence, wherein the heart rate change characteristics include instantaneous peak heart rate, duration of heart rate exceeding threshold, and rate of heart rate rise. Preset heart rate threshold range and corresponding light adjustment rules: When the instantaneous peak heart rate exceeds the upper limit of the heart rate threshold range, or the heart rate continues to exceed the threshold for a preset duration, or the heart rate rise rate exceeds the preset rate, the heart rate change characteristics are determined to meet the adjustment conditions. The first illumination adjustment command is generated based on the satisfied adjustment conditions: the greater the heart rate exceeds the threshold, the longer the duration of exceeding the threshold, or the faster the heart rate rises, the higher the increase ratio of light intensity and the greater the expansion angle of the illumination range in the first illumination adjustment command.

4. The intelligent headlight adjustment system for electric vehicles according to claim 1, characterized in that, The process involves determining road surface texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, identifying the road surface material based on the road surface texture features and ultrasonic reflection features, and generating a second illumination adjustment command for adjusting the illumination intensity based on the road surface material and the light absorption rate of different materials, including: The ultrasonic reflected signal is preprocessed by amplifying the weak signal and filtering out noise with a bandpass filter to extract ultrasonic reflection features, which include the reflected signal amplitude ratio, propagation time deviation and pulse width variation coefficient. The road surface image data is preprocessed by Gaussian filtering algorithm for noise reduction and histogram equalization algorithm for contrast enhancement. Convolutional neural network algorithm is used to extract the texture features of the road surface image. The texture features of the road surface image include local binary mode (LBP) histogram features and contrast and correlation features of gray-level co-occurrence matrix (GLCM). Construct a road surface material feature database, which stores ultrasonic reflection feature samples and road surface image texture feature samples corresponding to different road surface materials; The extracted ultrasonic reflection features and road surface image texture features are matched with samples in the road surface material feature database. The material corresponding to the sample with the highest matching degree is selected as the identified road surface material, and the light absorption rate parameter corresponding to the road surface material is retrieved. The expected road surface illumination intensity required for vehicle operation is preset. Based on the light absorption rate parameter corresponding to the road surface material and the expected road surface illumination intensity, the target emitted light intensity to be output by the headlight module is determined, and the second illumination adjustment command is generated.

5. The intelligent headlight adjustment system for electric vehicles according to claim 1, characterized in that, The fusion of vehicle speed signal and lateral acceleration data is used to determine whether the vehicle has entered a turning state using a decision tree algorithm, including: The vehicle speed signal and lateral acceleration data during the vehicle's movement are acquired, and the two types of data are processed for time synchronization and noise filtering. Construct a judgment feature system for the decision tree algorithm. The feature system includes key features related to vehicle turning behavior and corresponding judgment thresholds. The key features include at least vehicle speed features, vehicle lateral motion features, and feature state duration. Based on the aforementioned judgment feature system, a progressive decision tree judgment logic is established: first, the vehicle speed feature is used to determine whether the vehicle is in a speed range that is easy to turn; then, the vehicle body lateral movement feature is used to determine whether the vehicle body generates lateral movement related to turning; finally, the duration of the feature state is used to determine whether the above features constitute a stable turning-related state. The decision tree algorithm is used to comprehensively analyze the judgment results of the above features and output a judgment on whether the vehicle has entered a turning state. Based on the judgment result, the corresponding operation is triggered: if it is determined that the vehicle has entered a turning state, a trigger signal is generated to control the steering detection unit and tilt detection unit to switch from low power standby mode to high frequency working mode; if it is determined that the vehicle is not in a turning state, the low power standby mode of the steering detection unit and tilt detection unit is maintained.

6. The intelligent headlight adjustment system for electric vehicles according to claim 1, characterized in that, The process involves calculating the illumination angle adjustment value and illumination range expansion value based on preset fixed illumination area parameters, and generating a third illumination adjustment command for adjusting the illumination angle and illumination range based on the illumination angle adjustment value and the illumination range expansion value, including: Preset fixed lighting area parameters, wherein the fixed lighting area parameters include the ideal area length L0, width W0, and headlight installation height h of the area illuminated by the headlights when the vehicle is traveling straight; Receive and process the handlebar steering angle data θ collected by the steering detection unit and the vehicle body tilt angle data γ collected by the tilt detection unit; The illumination angle adjustment value is calculated based on the principle of geometric optics, including the horizontal turning angle adjustment value α=θ / 2 and the vertical tilt angle adjustment value δ=arctan(h / (L0cosγ)); The illumination range extension value β=θ is determined based on the handlebar steering angle θ, and the illumination range is extended towards the steering side; The horizontal steering angle adjustment value α, the vertical tilt angle adjustment value δ, and the illumination range expansion value β are written into the third illumination adjustment command, and the direction, magnitude, and range expansion direction and angle of the headlight angle adjustment are specified.

7. The intelligent headlight adjustment system for electric vehicles according to claim 1, characterized in that, The step of driving the vehicle headlight module to adjust the illumination angle, illumination range, and illumination intensity according to various illumination adjustment commands to achieve vehicle headlight adjustment includes: The control execution module is configured with an angle range adjustment unit and an intensity adjustment unit. The angle range adjustment unit is used to drive the vehicle lamp module to adjust the illumination angle and illumination range, and the intensity adjustment unit is used to adjust the illumination intensity of the vehicle lamp through current control. When the first illumination adjustment command is received, the angle range adjustment unit drives the headlight module to expand the illumination coverage range, and the intensity adjustment unit improves the illumination intensity of the headlight by smoothly adjusting the driving current of the headlight source. When the second illumination adjustment command is received, the intensity adjustment unit calculates the difference between the current illumination intensity and the target illumination intensity based on the target emitted illumination intensity determined in the second illumination adjustment command, and gradually adjusts the illumination intensity to the target value through a closed-loop control strategy. When the third illumination adjustment command is received, the angle range adjustment unit drives the headlight module to adjust the horizontal and vertical illumination angles, and at the same time drives the headlight module to expand the illumination range towards the turning side; if the angle and range adjustment causes a decrease in the illumination intensity per unit area, the intensity adjustment unit synchronously compensates and adjusts the illumination intensity. When multiple illumination adjustment commands are triggered simultaneously, the control execution module responds first to the third illumination adjustment command to complete the adjustment of the illumination angle and range; then, based on the adjusted angle and range, it superimposes the illumination intensity adjustment requirements from other illumination adjustment commands.

8. A method for intelligent headlight adjustment in electric vehicles, characterized in that, An intelligent headlight adjustment system for electric vehicles is provided, the system comprising a sensor module, a data processing module, a control execution module, and a headlight module, the method comprising: The sensor module acquires the driver's heart rate data, ultrasonic reflection signals from the road surface, road image data, and vehicle handlebar steering angle data and vehicle body tilt angle data. The data processing module determines the heart rate change characteristics based on the heart rate data, and generates a first light adjustment command for adjusting the light intensity and light range based on the heart rate change characteristics. The data processing module determines the road surface image texture features and ultrasonic reflection features based on the ultrasonic reflection signal and the road surface image data, identifies the road surface material based on the road surface image texture features and ultrasonic reflection features, and generates a second light adjustment command for adjusting the light intensity based on the road surface material and the light absorption rate of different materials. The data processing module integrates the vehicle speed signal and the vehicle's lateral acceleration data, and uses a decision tree algorithm to determine whether the vehicle has entered a turning state. If it is determined to be turning, the steering detection unit and the tilt detection unit are triggered to switch from low-power standby mode to high-frequency working mode, receive and process steering angle data and tilt angle data, and calculate the illumination angle adjustment value and illumination range expansion value in combination with preset fixed illumination area parameters. Based on the illumination angle adjustment value and the illumination range expansion value, a third illumination adjustment command is generated to adjust the illumination angle and illumination range. The control execution module drives the vehicle headlight module to adjust the illumination angle, illumination range and illumination intensity according to each illumination adjustment command, so as to realize the vehicle headlight adjustment.

9. A computer device, characterized in that, include: The system includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, the steps of the intelligent headlight adjustment method for electric vehicles as described in claim 8 are performed.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the intelligent headlight adjustment method for an electric vehicle as described in claim 8.

Images