A method and system for adaptive taillight illumination based on environmental perception

Abstract

This invention relates to the field of automotive electronic control technology and discloses a method and system for adaptive taillight illumination that integrates environmental perception. It aims to solve the problems of existing taillight systems relying on a single signal trigger, lacking environmental context awareness, and having rigid response strategies that hinder mass production. The method includes real-time acquisition of light intensity, millimeter-wave radar, camera, and meteorological data; construction of a dynamic environmental context model including visibility level, traffic complexity index, and emergency braking risk score; generation of response strategies for illumination area, brightness, flashing frequency, and duration based on a multi-dimensional mapping rule base; control of LED array partitions to perform adaptive illumination; and fine-tuning of trigger parameters through verification via a rear camera. The system includes multi-source sensors, an environmental modeling module, a strategy generator, an LED driving unit, and a feedback optimization module. Through this solution, the warning effectiveness of taillights, the precision and low latency of control, continuous learning capabilities, and mass production feasibility are synergistically improved.

Description

Technical Field

[0001] This invention relates to the field of automotive electronic control technology, specifically to a method and system for adaptive taillight illumination that integrates environmental perception. Background Technology

[0002] With the continuous evolution of intelligent vehicle technology, vehicle lighting systems have gradually developed from traditional passive signaling devices towards intelligence, environmental perception, and active interaction. As a key component for rear-end safety interaction, taillights not only perform basic signal transmission functions such as braking and steering, but also play an increasingly important role in improving driving safety, enhancing human-vehicle collaboration, and adapting to complex road environments.

[0003] Among them, the adaptive taillight illumination technology that integrates environmental perception aims to dynamically adjust the taillight illumination logic and display mode by acquiring real-time information about the vehicle's surrounding environment. This allows for a more accurate reflection of the vehicle's status and enhances the warning effect for road users behind. This technology typically relies on onboard sensors to comprehensively perceive multi-dimensional environmental parameters such as light intensity, weather conditions, traffic density, and distance to nearby vehicles, and optimizes the taillight response strategy accordingly.

[0004] Existing technologies for adaptive taillight control still have significant shortcomings: First, most systems trigger taillight illumination based on a single signal source (such as brake pedal action), lacking comprehensive judgment of the environmental context, resulting in insufficient warning effects in low visibility or complex traffic scenarios; second, existing solutions generally lack a dynamic mapping mechanism between environmental perception and taillight response, failing to intelligently adjust brightness, flashing frequency, or illuminated area according to different scenarios such as rain, fog, nighttime, or traffic congestion; third, some solutions attempting to introduce cameras or radar suffer from high computational latency, high power consumption, or high costs, making them difficult to popularize in mass-produced vehicles; finally, current taillight control systems are mostly closed designs, lacking data collaboration with other vehicle perception modules, limiting their integration potential in advanced driver assistance systems. These deficiencies create significant bottlenecks in improving active safety and human-centric interaction experiences for existing taillight systems, necessitating an efficient, low-cost, and environmentally adaptable adaptive illumination method and system architecture. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for adaptive lighting of automotive taillights that integrates environmental perception, which can effectively solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the technical solution adopted by this invention is as follows: a method and system for adaptive taillight illumination based on environmental perception, comprising the following specific steps: real-time acquisition of multi-source environmental perception data: synchronously acquiring the light intensity of the environment behind the vehicle, the relative distance and speed of nearby vehicles, traffic flow density, and current weather conditions through an onboard light intensity sensor, millimeter-wave radar, camera, and meteorological information module; constructing a dynamic environmental context model: performing time alignment and spatial registration on the multi-source environmental perception data, and fusing it to generate a three-dimensional environmental context vector containing visibility level, traffic complexity index, and emergency braking risk score; generating a taillight response strategy: based on the... The environmental context vector calls a preset multi-dimensional mapping rule library to determine the taillight's illumination area, brightness level, flashing frequency, and duration. This multi-dimensional mapping rule library is calibrated based on historical accident data and human factors engineering experiments. Adaptive lighting control is executed: the taillight response strategy is converted into driving signals to control the current intensity and on / off timing of each LED unit in the taillight array, achieving zoned, graded, and dynamic lighting effects. Feedback verification and strategy optimization: the actual taillight illumination status is monitored by a rear-facing camera and compared with the expected strategy. If the deviation exceeds a preset threshold, an online parameter fine-tuning mechanism is triggered to update the local weight coefficients in the multi-dimensional mapping rule library.

[0007] Preferably, in the real-time acquisition of multi-source environmental perception data, the sampling frequency of the light intensity sensor is not lower than a preset frequency threshold, and the measurement range covers a predetermined light intensity range, which is used to accurately distinguish lighting scenarios such as dusk, night, tunnels, and strong sunlight; the millimeter-wave radar operates in a preset high-frequency band, the detection distance range is within a predetermined working distance range, the distance resolution is a preset distance resolution, and the speed resolution is a preset speed resolution, which is used to track the approaching situation of vehicles behind in real time.

[0008] Preferably, the visibility level in the constructed dynamic environment context model is divided into multiple levels, where the lowest level corresponds to visibility higher than a preset high visibility threshold, and the highest level corresponds to visibility lower than a preset low visibility threshold; the traffic complexity index is calculated by weighting the number of adjacent vehicles, the standard deviation of relative speed, and the frequency of lane changes, and the value is within a preset numerical range; the emergency braking risk score is based on a comprehensive evaluation of the vehicle's deceleration, the following distance of the vehicle behind, and the road surface adhesion coefficient, and when the score is higher than a preset risk score threshold, it is determined to be a high-risk braking event.

[0009] Preferably, the multi-dimensional mapping rule base in the taillight response generation strategy is stored in non-volatile memory and contains a predetermined number of environment-response mapping entries. Each entry defines the taillight lighting parameters corresponding to a specific combination of visibility, traffic complexity, and risk score. When the environmental context vector falls into the coverage area of ​​multiple entries, a trilinear interpolation algorithm is used to generate an intermediate response strategy to ensure control continuity.

[0010] Preferably, in the adaptive illumination control, the taillight array is composed of a predetermined number or more independently controllable LED units, each unit supports multi-level brightness adjustment, and the minimum response delay is less than a preset time threshold; the illumination area is divided into an upper warning area, a middle braking area, and a lower contour area, wherein the upper warning area flashes at a preset flashing frequency during high-risk braking events, the brightness of the middle braking area increases linearly with the braking intensity, and the lower contour area remains constantly lit in low visibility environments to enhance vehicle contour recognition.

[0011] Preferably, in the feedback verification and strategy optimization, the frame rate of the rear camera is not lower than a preset frame rate threshold, and the resolution is not lower than a preset resolution threshold. The Euclidean distance between the actual lighting state and the expected strategy is calculated by analyzing the image brightness histogram and matching it with a preset lighting mode template. When the distance exceeds a preset deviation threshold for multiple consecutive frames, an online parameter fine-tuning mechanism is activated. The gradient descent method is used to adjust the weight coefficients of relevant entries in the multidimensional mapping rule base, and the learning rate is set to a preset learning rate parameter.

[0012] Preferably, the system communicates with the vehicle's electronic control unit via a CAN FD bus, with the communication rate within a preset high-speed communication range, ensuring low-latency interaction between environmental perception data and control commands; simultaneously, it receives forward collision warning signals from the advanced driver assistance system, and when a forward emergency braking command is received, it activates the taillight high warning mode in advance, shortening the response time to within a preset response time threshold.

[0013] Preferably, the multidimensional mapping rule base supports remote OTA updates and can receive optimized rule sets generated by cloud big data analysis through the vehicle network platform. The update cycle is no longer than a preset update cycle threshold. The cloud rule set is trained based on massive amounts of real road scene data and covers a variety of typical working conditions such as rain and fog, ice and snow, urban congestion, and highways.

[0014] Preferably, the taillight adaptive illumination method sets a low-power monitoring mode in the vehicle power management module. When the vehicle is parked and the ambient light is below a preset light threshold, only the light intensity sensor and millimeter-wave radar operate at minimum power. Once a moving object is detected approaching from behind, the entire system is immediately activated, with a wake-up delay less than a preset wake-up time threshold.

[0015] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0016] 1. Comprehensively improve the warning effectiveness of taillights

[0017] By integrating multi-source environmental perception data to construct a dynamic context model, this invention overcomes the limitations of traditional single braking signal triggering, enabling the taillight illumination logic to be highly matched with actual road risks. In low visibility or high traffic density scenarios, the taillights can automatically enhance brightness, expand the illuminated area, or introduce a flashing mode, significantly improving the recognition speed and reaction accuracy of drivers behind. Real-world test data shows that in adverse weather conditions, the braking reaction time of following vehicles is shortened by an average of more than a preset time.

[0018] 2. Achieve refined, low-latency adaptive control

[0019] Based on a high-resolution LED array and a millisecond-level response driving circuit, this invention supports independent control of different areas of the taillights. Combined with a multi-dimensional mapping rule base and interpolation algorithms, it can smoothly transition lighting strategies in complex environmental changes, avoiding abrupt switching. Simultaneously, through deep integration with the ADAS system via the CAN FD bus, it achieves feedforward control of forward-facing risk events, compressing taillight response latency to an industry-leading level.

[0020] 3. Possesses continuous learning and cloud-based collaboration capabilities.

[0021] By introducing feedback verification and online parameter fine-tuning mechanisms, the system can adapt to long-term changes such as device aging and environmental drift. Combined with OTA remote update function, it can continuously absorb massive amounts of real-world scenario data to optimize control strategies, forming a closed-loop intelligent system of "perception-decision-execution-learning" to ensure that system performance continues to improve rather than decline over time.

[0022] 4. Balancing high performance and mass production feasibility

[0023] This invention employs a mature and reliable combination of vehicle sensors, avoiding reliance on high-cost LiDAR or complex vision algorithms, thus keeping the overall hardware cost increase within a predetermined range. At the same time, through a low-power monitoring mode and efficient data processing flow, the system's average power consumption is lower than a preset power consumption threshold, meeting the electrical architecture requirements of mainstream vehicle models and possessing the basic conditions for large-scale mass production applications. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the overall technical solution architecture of the method and system of the present invention;

[0025] Figure 2 This is a schematic diagram of the core principle framework for the generation of taillight response strategies driven by the dynamic environment context model construction and multi-dimensional mapping rule base in this invention.

[0026] Figure 3 This is a flowchart illustrating the logical process framework for multi-source environmental perception data acquisition, fusion, and environmental context vector generation in this invention.

[0027] Figure 4 This is a schematic diagram of the closed-loop control interaction and data flow process of taillight adaptive lighting execution, status feedback verification and online parameter fine-tuning in this invention. Detailed Implementation

[0028] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] In a nighttime urban highway scenario, a passenger vehicle equipped with the adaptive taillight illumination system based on environmental perception described in this invention is traveling at 90 km / h. At this moment, the vehicle enters the exit area of ​​a long tunnel without streetlights, causing a sudden change in ambient light intensity. Simultaneously, two vehicles are approaching at high speed from behind, and the current weather is light fog with visibility of approximately 80 meters. Under this typical high-risk complex scenario, the system activates its full-process adaptive control mechanism.

[0030] First, real-time acquisition of multi-source environmental perception data is performed. The vehicle-mounted light intensity sensor continuously monitors the light intensity in the hemisphere behind the vehicle at a sampling frequency of 200 Hz, covering a range from 0.1 lux to 100,000 lux. It can accurately distinguish typical scenarios such as tunnels (approximately 50 lux), dusk (approximately 300 lux), nighttime urban roads (approximately 10 lux), and strong sunlight (>50,000 lux). Simultaneously, a 77 GHz millimeter-wave radar installed in the center of the rear bumper scans targets within a 150-meter range at a rate of 30 frames per second, with a distance resolution of 0.2 meters and a velocity resolution of 0.1 meters per second. It outputs the real-time relative distances of two vehicles behind at 45 meters and 78 meters, respectively, and their relative approach speeds at 12 meters per second and 5 meters per second, respectively. The front and side cameras, through image semantic segmentation algorithms, identify three vehicles in the current lane and adjacent lanes, one of which is performing a lane-changing maneuver. The meteorological information module obtains real-time meteorological data of the current location from the cloud meteorological service platform through the vehicle-mounted T-Box, confirming that the current situation is light haze, relative humidity is 85%, road surface temperature is 12 degrees Celsius, and there is no precipitation.

[0031] The system then proceeds to construct a dynamic environmental context model. The main control unit first timestamps the multi-source data, using hardware synchronization signals to ensure that the time deviation of all sensor data is less than 2 milliseconds. Next, spatial registration is performed, transforming the millimeter-wave radar point cloud coordinate system, camera image coordinate system, and vehicle body coordinate system to the rear field-of-view center coordinate system using a pre-calibrated extrinsic parameter matrix. Based on this, three core environmental indicators are calculated: visibility level, traffic complexity index, and emergency braking risk score.

[0032] Visibility is classified into five levels, with level 1 corresponding to visibility >500 meters, level 2 to 200-500 meters, level 3 to 100-200 meters, level 4 to 50-100 meters, and level 5 to <50 meters. Based on a combination of meteorological data and light intensity sensor readings, the current visibility level is determined to be level 4. Traffic Complexity Index The result is obtained from formula (1):

[0033]

[0034] in, This represents the number of nearby vehicles within a 50-meter radius behind, and its value is 2. The standard deviation of the relative speeds of these vehicles was calculated to be 4.2 m / s; Lane change frequency per minute, currently 1 time / minute; weighting coefficient =0.4、 =0.35、 =0.25, after normalization =0.78, falling within the [0,1] interval, indicating that the traffic flow is in a medium-to-high complexity state. The emergency braking risk score R is based on the vehicle's current deceleration. (If the brakes are not applied) =0), minimum following distance of the following vehicle Based on a comprehensive evaluation of (45 meters) and the road surface adhesion coefficient μ (0.65 obtained from meteorological data), when It was initially deemed high-risk. Although braking has not yet been initiated, due to... Smaller and with low visibility, The value was calculated to be 0.72, triggering a high-risk warning.

[0035] This generates a three-dimensional environment context vector. = [4, 0.78, 0.72].

[0036] Entering the taillight response strategy generation: The system loads a pre-stored multidimensional mapping rule base from non-volatile memory. This rule base stores over 100,000 environment-response mapping entries, each defined within a specific context. The combined taillight parameters include the illuminated area mask, brightness level (1-10), flashing frequency (0-5 Hz), and duration (0.5-5 seconds). Due to the current vector... If no entry is perfectly matched, the system uses a trilinear interpolation algorithm to interpolate among the eight nearest neighbor entries in the rule base, generating a continuous and smooth response strategy. The interpolation results are determined as follows: the upper warning zone is activated, flashing at a frequency of 3 Hz with a brightness level of 9; the middle braking zone remains constantly lit with a brightness level of 6 (only for preventative enhancement as no actual braking is applied); the lower contour zone remains constantly lit with a brightness level of 7 to enhance the overall vehicle contour.

[0037] Entering Adaptive Illumination Control: Adaptive illumination control is executed. The taillight array consists of 128 independently controllable LED units, physically divided into an upper warning zone (32 units), a central braking zone (64 units), and a lower contour zone (32 units). Each LED unit supports 10 levels of PWM dimming with a minimum response delay of 0.8 milliseconds. The drive circuit receives strategy commands from the main control unit, converts the brightness level into a corresponding duty cycle signal (e.g., level 9 corresponds to a 90% duty cycle), and precisely controls the current of each unit through a constant current drive chip. The 32 LEDs in the upper warning zone are switched on and off using a 3Hz square wave, creating a highly recognizable strobe effect; the central and lower areas maintain stable high brightness. The entire illumination process is completed within 1.5 milliseconds after receiving the strategy command, meeting real-time requirements.

[0038] Simultaneously, the system communicates with the vehicle's electronic control unit via the CAN FD bus at a rate of 8 megabits per second, synchronously receiving forward collision warning signals from the ADAS system. If a sudden obstacle ahead causes emergency braking, the ADAS will send a "high-risk forward" flag, and the system will immediately skip some environmental assessment steps and directly activate the highest warning mode (5 Hz flashing in the upper area, 10 brightness levels across the entire area), compressing the response time to less than 8 milliseconds.

[0039] The feedback verification and strategy optimization process begins. The rear camera, integrated above the license plate holder, continuously captures images of the taillights at 60 frames per second and a resolution of 1280×720. The image processing module extracts the brightness histogram of the taillight area and matches it against the expected lighting mode template (i.e., the strategy that generates the taillight response strategy output). The Euclidean distance D between the actual lighting state and the expected strategy is calculated. If D exceeds the threshold of 0.15 (normalized value) for five consecutive frames, a deviation is considered. This deviation may stem from LED aging, lens contamination, or driver circuit drift. In this case, the system initiates an online parameter fine-tuning mechanism, using gradient descent to adjust the weight coefficients of entries in the multidimensional mapping rule base that are adjacent to the current environment context vector V, with a fixed learning rate of 0.005. For example, if the actual brightness of the upper area is detected to be only 85% of the expected level, the target brightness level of that area under similar conditions is automatically increased, achieving closed-loop self-correction.

[0040] Furthermore, the system enters a low-power monitoring mode when the vehicle is parked. The power management module cuts off power to the taillight drive circuit and most modules of the main control unit, maintaining only the light intensity sensor and millimeter-wave radar operating at a low frequency of 10 Hz, with total power consumption below 0.5 watts. When the millimeter-wave radar detects a moving object approaching within 50 meters at a speed greater than 1 m / s and with a light intensity less than 50 lux, the system wakes up within 150 milliseconds and enters working mode, ensuring safety warnings when parking at night.

[0041] Example 2

[0042] In a rainy night scenario on a highway, a vehicle is cruising at 110 km / h when it encounters an accident ahead, requiring emergency braking. The core difference between this scenario and Example 1 is that the triggering mechanism shifts from being driven by environmental risk to being driven by ADAS feedforward control, and the taillight response strategy needs to be deeply coupled with braking intensity. This example focuses on the refined mapping logic of the multi-dimensional mapping rule base in generating the taillight response strategy under highly dynamic braking events, as well as the implementation of the linear mapping between brightness and braking intensity in the adaptive lighting control.

[0043] When the vehicle's brake pedal is depressed, the brake master cylinder pressure sensor outputs a pressure value of P = 8 MPa, corresponding to a deceleration of a = 6.2 m / s². Simultaneously, the ADAS system detects a stationary obstacle 60 meters ahead and issues an "emergency braking" command. The system immediately enters high-priority response mode.

[0044] In the real-time acquisition of multi-source environmental perception data, in addition to conventional environmental data, a braking intensity signal is introduced as a key input. The light intensity sensor reading is 8 lux (nighttime), the millimeter-wave radar detects the nearest vehicle behind as being 65 meters away with a relative speed of 8 meters per second, the weather module reports moderate rain, and visibility is approximately 120 meters.

[0045] In constructing the dynamic environmental context model, the visibility level was determined to be level 3, and the traffic complexity index T=0.65 (few vehicles behind). However, the emergency braking risk score R was calculated to be 0.89 due to high deceleration and slippery road surface (μ=0.45), which far exceeds the threshold of 0.7.

[0046] In the taillight response strategy, the multi-dimensional mapping rule base has a dedicated sub-rule set for high-risk braking events with R > 0.85. This sub-rule set quantifies the braking intensity 'a' into 5 levels (0-2, 2-4, 4-6, 6-8, >8 m / s²), and defines a different brightness level for the central braking zone for each level. The specific mapping relationship is: brightness level L = floor(a / 2) + 3, with an upper limit of 10. Currently, a = 6.2, so L = 6. Simultaneously, the upper warning zone is forced to flash at 4 Hz for a duration equal to the braking duration plus 1 second. The brightness of the lower contour area is fixed at level 8 to penetrate the rain.

[0047] In adaptive illumination control, the drive circuit receives the 'a' value from the braking system in real time and dynamically adjusts the PWM duty cycle of the central LED zone using a lookup table and linear interpolation. For example, as 'a' linearly increases from 0 to 6.2, the brightness of the central LED zone smoothly increases from level 3 to level 6 with a delay of less than 5 milliseconds, avoiding visual discomfort to drivers behind due to sudden brightness jumps. The upper LED zone, on the other hand, initiates a 4Hz flashing instantaneously when 'a' exceeds 4 m / s², creating a strong "emergency" visual signal.

[0048] During the feedback verification process, if the rear camera image becomes blurry due to rainwater, the system automatically switches to an indirect verification mode based on millimeter-wave radar echo intensity. This mode analyzes the increase in radar echo signal-to-noise ratio after the taillights are turned on to infer whether the actual luminous intensity of the taillights meets the standard. If the signal-to-noise ratio increase is insufficient, parameter fine-tuning is triggered.

[0049] This embodiment demonstrates how the present invention deeply integrates vehicle dynamics with environmental perception in active safety events, achieving precise synchronization between taillight response and driving behavior, which is impossible for traditional taillight systems that rely solely on brake switches.

[0050] Example 3

[0051] In the scenario of urban traffic congestion during morning rush hour, vehicles frequently start and stop, ambient light is overcast (approximately 500 lux), visibility is good (>500 meters), but traffic complexity is extremely high. The substantial difference between this scenario and the previous two examples lies in the fact that the taillight control strategy focuses on suppressing false triggers and reducing visual interference, rather than enhancing warnings. This embodiment focuses on the "degradation" strategy of the multi-dimensional mapping rule base in low-risk, high-complexity environments, and the extended application of low-power monitoring mode in urban parking scenarios.

[0052] Real-time acquisition of multi-source environmental sensing data shows: light intensity 500 lux, 6 vehicles within 50 meters behind, relative speed standard deviation. =1.8 m / s, lane change frequency =3 times / minute, no precipitation, road surface dry.

[0053] The dynamic environmental context model calculates visibility level 1. =0.92 (high complexity). =0.35 (Low risk, due to vehicle speed <10 km / h and no emergency braking).

[0054] In the taillight response generation strategy, the multidimensional mapping rule base is targeted at... >0.9 and In scenarios with a value of <0.5, the "anti-interference" strategy is enabled: the upper warning area is disabled (flickering frequency = 0), the middle braking area is only illuminated during actual braking, and the upper limit of the brightness level is set to 5 (to avoid frequent high brightness causing annoyance to drivers behind), and the lower outline area is turned off during the day (to save energy).

[0055] In adaptive illumination control, the taillight array only illuminates the central area when the brake pedal is applied, with brightness adjusted between levels 2 and 5 based on braking intensity. During non-braking periods, all taillights are off, consistent with traditional taillight behavior, but the system maintains environmental awareness internally to instantly switch to high-alert mode in case of sudden risks (such as a rapidly approaching vehicle).

[0056] Furthermore, when the vehicle is parked on the side of the road waiting for passengers, the system enters a low-power monitoring mode. Since the ambient light sensor detects >100 lux, only the millimeter-wave radar operates at 5 Hz. When a pedestrian or bicycle is detected approaching within 10 meters at a speed >0.5 m / s, the system wakes up and illuminates the lower contour area (brightness level 3) for 10 seconds, providing a gentle contour warning rather than a glaring brake light, balancing safety and the needs of civilized urban driving.

[0057] This embodiment demonstrates the intelligent "restraint" capability of the present invention in non-dangerous scenarios, avoiding light pollution and driver fatigue caused by excessive warnings, which is an environmental intelligent adaptation feature not possessed by existing technologies.

[0058] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

[0059] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for adaptive illumination of automotive taillights based on environmental perception, characterized in that, Includes the following steps: Real-time acquisition of multi-source environmental perception data, including rear illumination intensity obtained by vehicle-mounted light intensity sensor, relative distance and speed of nearby vehicles obtained by millimeter-wave radar, traffic flow density information obtained by camera, and current weather conditions obtained by meteorological information module. The multi-source environmental perception data is time-aligned and spatially registered, and then fused to generate a three-dimensional environmental context vector containing visibility level, traffic complexity index and emergency braking risk score. Based on the environmental context vector, a preset multidimensional mapping rule library is invoked to determine the taillight illumination area, brightness level, flashing frequency, and duration. The multidimensional mapping rule library stores taillight illumination parameters corresponding to a specific combination of visibility level, traffic complexity index, and emergency braking risk score. The taillight response strategy is converted into a driving signal to control the current intensity and on / off timing of each LED unit in the taillight array, thereby achieving a zoned, hierarchical, and dynamic lighting effect. The actual lighting status of the taillights is monitored by the rear camera and compared with the expected strategy. If the deviation exceeds the preset threshold, an online parameter fine-tuning mechanism is triggered to update the local weight coefficients in the multi-dimensional mapping rule base.

2. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, The visibility level is divided into multiple levels, with the lowest level corresponding to visibility above the high visibility threshold and the highest level corresponding to visibility below the low visibility threshold. The traffic complexity index is calculated by weighting the number of adjacent vehicles, the standard deviation of relative speed, and the frequency of lane changes. The emergency braking risk score is based on a comprehensive assessment of the vehicle's deceleration, the following distance of the vehicle behind, and the road surface adhesion coefficient.

3. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, When the environmental context vector falls within the coverage area of ​​multiple mapping entries, a trilinear interpolation algorithm is used to generate an intermediate response strategy to ensure the continuity of taillight control.

4. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, The taillight array is divided into an upper warning zone, a middle braking zone, and a lower contour zone. The upper warning zone flashes at a set flashing frequency when the emergency braking risk score is higher than the risk score threshold. The brightness of the middle braking zone increases linearly with the braking intensity. The lower contour zone remains constantly lit when the visibility level is higher than a set level.

5. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, By analyzing the image brightness histogram and matching it with the preset lighting mode template, the Euclidean distance between the actual lighting state and the expected strategy is calculated. When the distance exceeds the preset deviation threshold for multiple consecutive frames, an online parameter fine-tuning mechanism is activated, and the weight coefficients of relevant entries in the multidimensional mapping rule base are adjusted using the gradient descent method.

6. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, The system communicates with the vehicle's electronic control unit via the CAN FD bus and receives forward collision warning signals from the advanced driver assistance system; when it receives a forward emergency braking command, it activates the high warning mode of the taillights in advance.

7. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, The multidimensional mapping rule base supports remote OTA updates and receives optimized rule sets generated by cloud big data analysis through the vehicle networking platform. The optimized rule sets cover a variety of typical operating conditions, including rain and fog, ice and snow, urban congestion, and highways.

8. The adaptive taillight illumination method for automobiles based on integrated environmental perception as described in claim 1, characterized in that, When the vehicle is parked and the ambient light is below the preset light threshold, the system enters a low-power monitoring mode, maintaining only the light intensity sensor and millimeter-wave radar operation; when a moving object is detected approaching from behind, the entire system is immediately awakened and put into operation.

9. A vehicle taillight adaptive lighting system integrating environmental perception, characterized in that, include: The multi-source environmental sensing module is used to collect real-time data on light intensity, relative distance and speed of nearby vehicles, traffic flow density, and weather conditions. The environmental context modeling module is used to align and register multi-source environmental perception data in time and space to generate a three-dimensional environmental context vector containing visibility level, traffic complexity index and emergency braking risk score; the strategy generation module is used to call a multi-dimensional mapping rule base based on the environmental context vector to output the taillight illumination area, brightness level, flashing frequency and duration. The execution control module is used to convert the taillight response strategy into drive signals to control the current intensity and on / off timing of each LED unit in the taillight array. The feedback verification module is used to monitor the actual lighting status of the taillights through the rear camera, and triggers an online parameter fine-tuning mechanism to update the local weight coefficients in the multidimensional mapping rule base when the deviation exceeds a preset threshold.

10. The adaptive taillight illumination system for automobiles based on integrated environmental perception as described in claim 9, characterized in that, The taillight array consists of independently controllable LED units, divided into an upper warning zone, a middle braking zone, and a lower contour zone. The system communicates with the vehicle's electronic control unit via the CAN FD bus, receives forward collision warning signals from the advanced driver assistance system, and activates the high warning mode in advance when a forward emergency braking command is received. The multidimensional mapping rule base is stored in non-volatile memory, supports remote OTA updates, and enters a low-power listening mode when the vehicle is parked and the light intensity is below the threshold.

Images