A method for coordinated control of Micro-led digital headlights and AFS actuators

By constructing a collaborative control system between Micro-LED digital headlights and AFS actuators, the problems of unstable beam pointing and insufficient system safety in existing technologies have been solved, achieving high-precision and fast-response cornering lighting and anti-glare control, thus improving the overall performance and safety of the system.

CN123104514APending Publication Date: 2026-11-17RUIBO YIHAO (SHANGHAI) TECHNOLOGY DEVELOPMENT CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202611531352.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-30
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing Micro-LED digital headlight systems suffer from problems such as nonlinear errors, response lag, anti-glare shielding zone drift, frequent switching of lighting modes, and uncoordinated fault handling in cornering lighting control, resulting in unstable beam pointing and insufficient system safety.

Method used

A collaborative control system for Micro-LED digital headlights and AFS actuators is constructed. The MCU main controller integrates data acquisition, preprocessing, lighting mode management, AFS angle control and safety management modules to achieve multi-source information fusion, real-time beam deflection and graded safety degradation, ensuring beam accuracy and stability.

Benefits of technology

It improves the accuracy and responsiveness of AFS cornering lighting control, eliminates anti-glare shading zone drift, stabilizes lighting mode switching, enhances system safety and integration, and reduces costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123104514A_ABST
    Figure CN123104514A_ABST
Patent Text Reader

Abstract

A method for coordinated control of Micro-LED digital headlights and AFS actuators includes: a data acquisition module acquiring vehicle data vectors and sending them to an MCU main controller; a preprocessing module preprocessing the vehicle data vectors to obtain the filtered effective values ​​and validity flags of each data point; a lighting mode management module determining the lighting mode based on the filtered data and outputting the corresponding lighting control scheme; an AFS angle control module calculating the horizontal and vertical deflection angles of the headlights based on the steering wheel angle and vehicle speed, and generating motor execution commands; a Micro-LED beam control module outputting Micro-LED lighting drive commands based on the camera target list and the actual mechanical angle of the AFS; a safety management module generating safety level adjustment control commands based on the filtered effective values ​​and validity flags of each data point; and an execution module implementing coordinated control and drive of the Micro-LED digital headlights and AFS actuators based on the lighting control scheme, motor execution commands, Micro-LED lighting drive commands, and safety level adjustment control commands.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of automotive intelligent lighting control technology, and in particular to a method for coordinated control of a Micro-led digital headlight and an AFS actuator. Background Technology

[0002] With the development of intelligent and digital lighting technologies in automobiles, traditional LED headlights have gradually evolved towards ADB (Adaptive High Beam), AFS (Adaptive Front Lighting System), and Micro-LED digital headlights. Micro-LED digital headlights, with their high brightness, high resolution, fast response speed, and ability to achieve dynamic projection, can adjust the lighting area and projection content in real time according to the road environment, traffic participants, and vehicle operating status.

[0003] In existing technologies, Micro-LED digital headlight systems typically employ independent LED drive control units to drive the pixel light sources, and independent motor control modules to achieve horizontal / vertical adjustment of the AFS actuator. These functional modules are interconnected via CAN or LIN buses. At the control algorithm level, existing technologies have the following specific drawbacks:

[0004] (1) AFS cornering lighting control mostly adopts a single-input lookup table method: it only looks up the lamp body deflection angle based on the steering wheel angle through a fixed mapping table, without integrating multi-source information such as yaw rate and navigation aiming curvature. On the one hand, the steering wheel angle has a nonlinear error with the actual front wheel angle after passing through the steering transmission ratio, suspension roll and other factors, and the beam direction has a steady-state deviation in low-speed large-angle conditions (parking, U-turn) and high-speed gentle curve conditions (ramp, long curve); on the other hand, the lookup table method responds with a lag in the curve entry process, and the beam "chasing curve" delay is obvious.

[0005] (2) Vertical slope compensation is mostly open-loop control or only relies on IMU attitude angle: it cannot distinguish between static road slope (which requires continuous compensation) and dynamic pitch caused by braking / acceleration (which only requires transient suppression), resulting in glare for oncoming vehicles when going uphill or insufficient near-field illumination when going downhill: it also does not combine the suspension height sensor to perform joint calculation of load and slope.

[0006] (3) ADB pixel-level anti-glare and AFS mechanical deflection are independent of each other: the anti-glare shielding area is calculated in the image coordinate system and directly output to the pixel matrix without being corrected in real time with the mechanical posture of the lamp body; when the AFS mechanism rotates (the mechanical response is usually at the level of hundreds of milliseconds), the shielding area drifts in the road coordinate system, which poses a risk of brief glare to oncoming vehicles.

[0007] (4) The logic for determining the lighting mode is simple: lacking hysteresis and timing preservation mechanisms, the frequent switching of modes caused by the jitter of criteria such as vehicle speed and ambient light near the threshold results in light distribution jumps and visual flicker.

[0008] (5) The fault handling adopts a simple shutdown strategy: when faults such as motor stall, sensor failure, or communication interruption occur, the corresponding function is directly shut down. There is a lack of a graded degradation mechanism, and the rapid response capability of Micro-LED pixel-level control is not utilized to compensate for faults caused by mechanical actuator mismatch. Lighting control and AFS actuator control are independent of each other. There is a lack of a unified and coordinated control mechanism between digital light source light distribution and mechanical steering angle, resulting in response delay or matching error between the beam illumination area and the mechanical steering angle. To clarify the boundary between this invention and the closest prior art, the following representative prior art documents are further analyzed:

[0009] Chinese invention patent application CN118514600A discloses a method for adjusting high beams based on steering wheel angle control. It converts the steering wheel angle into the front wheel angle according to the steering transmission ratio, calculates the steering radius using a vehicle model with an understeering stability coefficient, and replaces motor adjustment with LED electronic control. However, it still relies on single-source geometric calculation of the steering wheel angle: it does not integrate yaw rate and navigation pre-aiming curvature, lacks a multi-source fusion mechanism that adapts to signal validity, and does not have two-level coordination between mechanical deflection and pixel matrix. Furthermore, it fails to address the issue of the anti-glare shielding area drifting with the mechanical rotation of the headlight body.

[0010] Therefore, there is an urgent need for a hierarchical modular collaborative control method for Micro-LED digital headlights and AFS actuators. This method should achieve unified collaborative control of mode management, cornering lighting, slope compensation, and pixel-level anti-glare through a clear algorithm chain of data acquisition, decision calculation, and command output, and should also have graded safety degradation capabilities. Summary of the Invention

[0011] The present invention provides a method for coordinated control of Micro-led digital headlights and AFS actuators, which can realize coordinated control of Micro-led digital headlights and AFS actuators.

[0012] To achieve the above objectives, the present invention provides a method for coordinated control of a Micro-led digital headlight and an AFS actuator, the key of which includes the following steps:

[0013] A method for coordinated control of a Micro-led digital headlight and an AFS actuator, characterized by the following steps:

[0014] Step 1: Construct a collaborative control system for Micro-led digital headlights and AFS actuators. This collaborative control system is equipped with an MCU main controller, which is connected to a data acquisition module and an execution module.

[0015] The MCU main controller integrates a preprocessing module, a lighting mode management module, an AFS angle control module, a Micro-LED beam control module, and a safety management module;

[0016] Step 2: The data acquisition module periodically acquires vehicle data vector x and transmits it to the MCU main controller;

[0017] Step 3: The preprocessing module in the MCU main controller performs preprocessing operations on the vehicle data vector x to obtain the filtered effective value and validity flag of each data.

[0018] Step 4a: The lighting mode management module determines the lighting mode based on the effective value and validity flag of each data, obtains the current vehicle lighting mode, and then outputs the corresponding lighting control scheme to the execution module according to the lighting mode.

[0019] Step 4b: The AFS angle control module determines the steering wheel angle based on the vehicle data vector x. Given the vehicle speed v, the system calculates the horizontal and vertical deflection angles of the headlights in real time and generates motor execution commands for the execution module.

[0020] Step 4c: The Micro-LED beam control module outputs a Micro-LED light driving command to the execution module based on the camera target list in the vehicle data vector x and the AFS mechanical actual angle;

[0021] Step 4d: The security management module generates a security level upgrade / downgrade control command for the execution module based on the filtered effective value and validity flag of each data.

[0022] Step 5: The execution module implements coordinated control and drive of the Micro-led digital headlights and AFS actuators based on the lighting control scheme, motor execution instructions, Micro-LED lighting drive instructions, and safety level adjustment control instructions.

[0023] Through the above design, this invention establishes a modular control architecture of perception layer-decision layer-execution layer by setting up a data acquisition module, MCU main controller and execution module, and clearly defines the input signals, algorithm operation relationships and output instructions of each module. This invention not only provides an AFS horizontal curve lighting algorithm based on multi-source information fusion and a vertical slope compensation algorithm with static and dynamic separation, so that the beam spreads out according to the curve and slope, but also provides a pixel-level anti-glare algorithm that is linked with AFS mechanical deflection in real time, so that the obscured area always locks onto traffic participants during the rotation of the light body. In addition, this invention eliminates mechanical response lag through a two-level beam deflection coordination mechanism of "mechanical coarse adjustment + pixel fine adjustment", and provides a lighting mode management algorithm with hysteresis and hold time and a graded failure degradation strategy to improve the functional safety level of the system.

[0024] Preferably, in step 2, the vehicle data vector x includes, but is not limited to, the vehicle's steering wheel angle. Vehicle speed v, yaw rate ω_z, longitudinal acceleration Wiper speed (G_w), headlight switch status (S_L), gear position signal (R_g), road type (RoadType), navigation aiming curvature (K_nav), camera target list, motor position feedback (P_fb), ambient light level. Lamp body temperature T_sen, channel current I_ch, channel voltage U_ch data, front suspension height h_f and rear suspension height h_r;

[0025] The data acquisition module is equipped with a vehicle environment bus, a task execution bus, and an ADC sampling circuit.

[0026] The MCU main controller is connected to the body controller and the intelligent driving domain controller via the vehicle environment bus, and is used to obtain the vehicle's steering wheel angle. Vehicle speed v, yaw rate ω_z, longitudinal acceleration Wiper speed (G_w), headlight switch status (S_L), gear position signal (R_g), road type (RoadType), navigation aiming curvature (K_nav), and camera target list data;

[0027] The MCU main controller is connected to the AFS actuator via a task execution bus to acquire the vehicle's motor position feedback P_fb data;

[0028] The MCU main controller obtains the ambient light level via an ADC sampling circuit. Lamp body temperature T_sen, channel current I_ch, and channel voltage U_ch data.

[0029] The MCU main controller periodically receives sensor data frames sent by the body controller and intelligent driving domain controller via the vehicle environment bus module, and simultaneously collects the internal status of the lamp body via the local ADC sampling circuit.

[0030] Preferably, the preprocessing module is provided with a limiting filter unit, a low-pass filter unit, and a validity flag and timeout monitoring unit connected in sequence, and a dead zone and normalization processing unit is also provided between the low-pass filter unit and the validity flag and timeout monitoring unit.

[0031] The amplitude limiting filter unit is used to perform amplitude limiting filtering on the original signal, and the low-pass filter unit is used to perform low-pass filtering on the amplitude-limited filtered signal. The dead-zone and normalization processing unit performs dead-zone processing on the steering wheel angle signal and logarithmically normalizes the ambient light intensity. The validity flag and timeout monitoring unit is used to monitor each signal in real time and assign values ​​to the validity flags of each signal.

[0032] Preferably, in step 4a, the lighting mode includes 8 lighting modes, namely: off / position light mode M0, low beam basic mode M1, city mode M2, highway mode M3, ADB adaptive high beam mode M4, cornering enhancement mode M5, bad weather mode M6, welcome / home mode M7 and reversing auxiliary lighting mode M8.

[0033] The output light mode at any given time is determined by the light mode that meets the entry conditions and has the highest priority. The light mode priorities are as follows:

[0034] P(M8)>P(M7)>P(M6)>P(M4)>P(M3)>P(M2)>P(M1)>P(M0);

[0035] The curve enhancement mode M5 is a parallel superposition mode. It does not participate in the above priority arbitration, but is superimposed on the current ground state modes M1~M4 as a modification layer.

[0036] The input data for the lighting mode management module includes: vehicle speed v, ambient light intensity E, light switch status S_L, wiper gear position G_w, gear position signal R_g, road type RoadType, number of valid targets from the camera N_obj, and steering wheel angle. The navigation aiming curvature K_nav, and the system power-on / power-off status.

[0037] Preferably, in step 4b, the AFS angle control module is provided with a horizontal control unit, a vertical control unit, a two-level collaborative unit, and a motor closed-loop unit;

[0038] The horizontal control unit is used for dynamic curve lighting using a horizontal curve lighting algorithm; the vertical control unit is used for slope compensation using a vertical slope compensation algorithm. The two-stage collaborative unit employs a two-stage beam deflection collaborative mechanism of "mechanical coarse adjustment + pixel fine adjustment" to perform overall translation of the pixel brightness matrix output by the beam control module. The motor closed-loop unit is used for motor closed-loop control and execution output.

[0039] Preferably, in step 4c, the Micro-LED beam control module takes the camera target list and the AFS mechanical actual angle as input and outputs the brightness setting value of each pixel in the Micro-LED pixel matrix.

[0040] Preferably, the safety management module is used to continuously monitor the system's operating status and perform temperature derating, power budget allocation, and graded failure degradation control.

[0041] Preferably, the execution module includes an LED driving circuit, a high-side driving circuit, and a motor driving circuit. The MCU main controller drives the MicroLED combination via the LED driving circuit, drives the high-side load via the high-side driving circuit, and drives the AFS execution motor via the motor driving circuit.

[0042] The beneficial effects of this invention are:

[0043] (1) Improved accuracy and responsiveness of AFS cornering lighting control. This invention achieves multi-source weighted fusion of steering wheel angle geometric curvature, yaw rate curvature, and navigation aiming curvature through a horizontal control unit. The weights are zoned according to vehicle speed and adaptively reset to zero according to signal validity, overcoming the steady-state deviation of the single steering wheel angle lookup method under low-speed large angle and high-speed gentle curve conditions. Combined with dead zone, speed adaptive amplitude limiting and rate limiting, it ensures that the beam pointing is stable and jitter-free.

[0044] (2) Improved slope compensation accuracy. By separating and filtering static slope and dynamic pitch using the vertical control unit, and combining it with longitudinal acceleration feedforward compensation, over-adjustment of transient pitch and insufficient compensation of static slope are avoided, effectively suppressing glare from the oncoming vehicle under braking pitch conditions.

[0045] (3) The anti-glare shielding area is linked with the mechanical deflection in real time. The target angle is transformed into the lamp body coordinates and then mapped to the pixel matrix through the Micro-LED beam control module. During the mechanical rotation, the shielding area remains locked in the road coordinate system, eliminating the risk of temporary glare caused by the drift of the ADB shielding area with the rotation of the lamp body in the prior art. In addition, the asymmetric shielding establishment / de-establishment sequence, namely immediate establishment, delayed maintenance and gradual release, effectively eliminates the light flicker caused by the detection of flicker.

[0046] (4) Two-stage beam deflection coordination eliminates mechanical lag. The two-stage coordination unit shortens the beam pointing response time from about 200ms mechanical response time to about 5ms one pixel refresh cycle through the "mechanical coarse adjustment + pixel fine adjustment" mechanism, which greatly shortens the beam pointing response time; and the mechanical angle undertakes the main deflection range without sacrificing the illumination field of view, thus improving the beam illumination quality.

[0047] (5) Stable mode management. The lighting mode determination adopts a three-layer arbitration structure: the first layer is the basic lamp group layer, determined by S_L; the second layer is the operating condition adaptation layer, selected from city / highway / severe weather / ADB; the third layer is the instantaneous event layer, such as reversing, overtaking, and welcoming. The three-layer arbitration structure, together with the hysteresis band and duration criteria, effectively eliminates the frequent switching of different lighting modes near the threshold, and the problem of light distribution jumps and flickering.

[0048] (6) Enhanced fail-safe capability. The safety management module retains functions in different scenarios such as sensor failure, camera failure, motor stall, and bus failure through a six-level degradation state machine. In particular, when the motor fails, electronic compensation is performed by Micro-LED pixel shifting, so that the cornering lighting function can still be partially used after mechanical failure. The power budget is allocated according to functional priority to ensure that the basic low beam distribution is not affected.

[0049] (7) System integration and cost advantages. All decision-making algorithms of this invention run on a single MCU main controller, reusing the dual-bus hardware architecture, in which the environment bus is responsible for sensing data access and the task bus is responsible for issuing and feeding back execution instructions. No additional controllers and wiring harnesses are required, which has the advantages of high system integration and low cost of use. Attached Figure Description

[0050] Figure 1 This is a flowchart of the method of the present invention;

[0051] Figure 2 This is a block diagram of the collaborative control system structure of Micro-led digital headlights and AFS actuator in the embodiment;

[0052] Figure 3 This is a schematic diagram of the masking template in the embodiment;

[0053] Figure 4 This is a schematic diagram of the AFS curve pre-aiming geometry in the embodiment. Detailed Implementation

[0054] The present invention will be further described in detail below with reference to the accompanying drawings and specific examples. The following embodiments or drawings are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0055] like Figure 1As shown, a method for coordinated control of a Micro-led digital headlight and an AFS actuator includes the following steps:

[0056] Step 1: Construct a collaborative control system for Micro-led digital headlights and AFS actuators. This collaborative control system is equipped with an MCU main controller, which is connected to a data acquisition module and an execution module.

[0057] The MCU main controller integrates a preprocessing module, a lighting mode management module, an AFS angle control module, a Micro-LED beam control module, and a safety management module;

[0058] Step 2: The data acquisition module periodically acquires vehicle data vector x and transmits it to the MCU main controller;

[0059] Step 3: The preprocessing module in the MCU main controller performs preprocessing operations on the vehicle data vector x to obtain the filtered effective value and validity flag of each data.

[0060] Step 4a: The lighting mode management module determines the lighting mode based on the effective value and validity flag of each data, obtains the current vehicle lighting mode, and then outputs the corresponding lighting control scheme to the execution module according to the lighting mode.

[0061] Step 4b: The AFS angle control module determines the steering wheel angle based on the vehicle data vector x. Given the vehicle speed v, the system calculates the horizontal and vertical deflection angles of the headlights in real time and generates motor execution commands for the execution module.

[0062] Step 4c: The Micro-LED beam control module outputs a Micro-LED light driving command to the execution module based on the camera target list in the vehicle data vector x and the AFS mechanical actual angle;

[0063] Step 4d: The security management module generates a security level upgrade / downgrade control command for the execution module based on the filtered effective value and validity flag of each data.

[0064] Step 5: The execution module implements coordinated control and drive of the Micro-led digital headlights and AFS actuators based on the lighting control scheme, motor execution instructions, Micro-LED lighting drive instructions, and safety level adjustment control instructions.

[0065] In step 2, the vehicle data vector x includes, but is not limited to, the vehicle's steering wheel angle. Vehicle speed v, yaw rate ω_z, longitudinal acceleration Wiper speed (G_w), headlight switch status (S_L), gear position signal (R_g), road type (RoadType), navigation aiming curvature (K_nav), camera target list, motor position feedback (P_fb), ambient light level. Lamp body temperature T_sen, channel current I_ch, channel voltage U_ch data, front suspension height h_f and rear suspension height h_r;

[0066] The data acquisition module is equipped with a vehicle environment bus, a task execution bus, and an ADC sampling circuit.

[0067] The MCU main controller is connected to the body controller and the intelligent driving domain controller via the vehicle environment bus, and is used to obtain the vehicle's steering wheel angle. Vehicle speed v, yaw rate ω_z, longitudinal acceleration Wiper speed (G_w), headlight switch status (S_L), gear position signal (R_g), road type (RoadType), navigation aiming curvature (K_nav), and camera target list data;

[0068] The MCU main controller is connected to the AFS actuator via a task execution bus to acquire the vehicle's motor position feedback P_fb data;

[0069] The MCU main controller obtains the ambient light level via an ADC sampling circuit. Lamp body temperature T_sen, channel current I_ch, and channel voltage U_ch data.

[0070] In this embodiment, the data flow between each layer is a unidirectional main link plus a feedback branch: the perception layer outputs to each module of the decision layer; the output of each module of the decision layer is arbitrated by the security management module and then sent to the execution layer; the position feedback and current and voltage sampling of the execution layer are transmitted back to the decision layer through the perception layer to form a closed loop.

[0071] The system control cycle and timing conventions are shown in Table 1. These can be adjusted according to the vehicle model calibration.

[0072] Table 1 System Control Cycle and Timing Conventions

[0073]

[0074] The preprocessing module is provided with a limiting filter unit, a low-pass filter unit, and a validity flag and timeout monitoring unit connected in sequence. A dead zone and normalization processing unit is also provided between the low-pass filter unit and the validity flag and timeout monitoring unit.

[0075] In step 3, the preprocessing module processes each raw signal in the vehicle data vector x. The preprocessing operation is performed, and the specific steps are as follows:

[0076] Step S1: The amplitude limiting filter unit processes each original signal acquired cycle by cycle. Perform amplitude limiting filtering if the current frame signal Compared with the previous valid frame signal The difference is greater than the limit of the maximum rate of change of the signal in a single period. Then the current frame signal Frames identified as abnormal are discarded, and the previous valid frame is replaced. As the current valid frame signal Output the signal; otherwise, output the current frame signal. As the current valid frame signal To output, the logical expression is:

[0077] when At that time, ;

[0078] otherwise, ;

[0079] Where k represents the sampling frame index and i represents the original signal type index;

[0080] Step S2: The low-pass filter unit processes the current valid frame signal that has passed through the amplitude limiting filter unit. First-order inertial low-pass filtering is performed to suppress high-frequency noise, and the effective values ​​of the filtered signals are obtained. The expression is:

[0081] ;

[0082] ;

[0083] in, The effective value of the filter in the k-th frame of the original signal of type i is given. For signal acquisition period, This is the filtering time constant; when the original signal is the steering wheel angle, =50ms; when the original signal is vehicle speed. =100ms; when the original signal is the yaw rate. =80ms; when the original signal is the suspension height. =200ms; when the original signal is illuminance. 500ms.

[0084] Step S3: The dead zone and normalization processing unit filters the effective value of the steering wheel angle in the original signal. To perform dead-zone handling and eliminate slight hand tremors in the driver, the expression is:

[0085] when hour, ;

[0086] otherwise, ;

[0087] in, This is the effective value of the filtered steering wheel angle. The steering wheel angle after dead zone processing. The dead zone steering wheel angle threshold. For the sign function; steering wheel angle data after dead zone processing. It is only used for AFS horizontal curve lighting calculation and does not change other control logic of the vehicle steering system.

[0088] The dead zone and normalization processing unit performs logarithmic normalization on the effective value E of the ambient light intensity in the original signal to match the human eye response. The expression is as follows:

[0089] ;

[0090] in, The normalized ambient illuminance This represents the effective value of the filtered ambient light intensity. For normalized reference illuminance parameters, ;

[0091] This normalization compresses the wide range of 0~200001x into a logarithmic interval of 0~2.3. The entry threshold of 5001x for the lighting mode hysteresis mechanism corresponds to... ≈0.778. Exit threshold 700 Ix ≈0.903. (Normalized value) It is sent to the lighting mode management module for ambient light comparison in the three-layer arbitration.

[0092] Step S4: The validity flag and timeout monitoring unit monitor each signal in real time. When a monitored signal meets any of the following conditions, the validity flag of that signal is set to invalid, Valid=0:

[0093] (1) The number of consecutive dropped frames of the signal is greater than or equal to the set frame number threshold. ;

[0094] (2) Communication timeout: No update for more than 3 data collection cycles;

[0095] (3) Exceeding the physical range: The signal value exceeds the range specified in the table above;

[0096] (4) Verification error: CAN message CRC check failed or ADC self-test abnormal. Sensor fault flag: Internal fault bit from the sensor is set.

[0097] Otherwise, the signal is marked as valid, Valid=1;

[0098] Step S5: The preprocessing module finally outputs the filtered effective values ​​of each signal and their effectiveness flags, as well as the steering wheel angle after dead zone processing. Normalized ambient illuminance .

[0099] In step 4a, the lighting mode includes 8 lighting modes, namely: off / position light mode M0, low beam basic mode M1, city mode M2, highway mode M3, ADB adaptive high beam mode M4, cornering enhancement mode M5, inclement weather mode M6, welcome / home mode M7 and reversing auxiliary lighting mode M8.

[0100] The mode determination adopts a "three-layer arbitration" structure: the first layer is the basic light group layer (determined by SL), the second layer is the operating condition adaptation layer (selected from city / highway / severe weather / ADB), and the third layer is the instantaneous event layer (reversing, overtaking, welcoming). The output light mode at any given time is determined by the light mode that meets the entry conditions and has the highest priority. The light mode priorities are as follows:

[0101] P(M8)>P(M7)>P(M6)>P(M4)>P(M3)>P(M2)>P(M1)>P(M0);

[0102] The curve enhancement mode M5 is a parallel superposition mode. It does not participate in the above priority arbitration, but is superimposed on the current ground state modes M1~M4 as a modification layer.

[0103] The input data for the lighting mode management module includes: vehicle speed v, ambient light intensity E, light switch status S_L, wiper gear position G_w, gear position signal R_g, road type RoadType, number of valid targets from the camera N_obj, and steering wheel angle. The navigation aiming curvature K_nav, and the system power-on / power-off status.

[0104] The correspondence between each decision input and the affected mode is shown in Table 2:

[0105] Table 2 shows the correspondence between each judgment input and the affected mode.

[0106]

[0107] The lighting mode determination logic of the lighting mode management module is as follows:

[0108] (1) Automatic headlight activation: A prerequisite for entering the low beam basic mode M1 is that the headlight switch is in automatic mode, i.e., when the headlight switch state S_L=AUTO:

[0109] When E<E_on=500Lx, set the day-night flag DayNight=night;

[0110] When E>E_off=700Lx and the duration reaches the set illumination time threshold T_E, where T_E=5s, set the day-night flag DayNight=day;

[0111] The 200Lx hysteresis band gap between E_on and E_off is used to prevent repeated switching under dusk / tree shadow working conditions.

[0112] Wherein, E represents the ambient illuminance, E_on represents the automatic lighting turn-on threshold, and E_off represents the automatic lighting turn-off threshold;

[0113] (2) Urban mode M2:

[0114] Entry condition: Vehicle speed v<vehicle speed threshold v_city, v_city=50km / h, and the duration reaches the set speed time threshold T_hold, T_hold=10s, and the road type RoadType is urban area or the number of valid camera targets N_obj≥the set target number threshold, and the day-night flag DayNight=night;

[0115] Exit condition: Vehicle speed v>vehicle speed threshold v_city+hysteresis vehicle speed v_0 or the road type RoadType changes to expressway; v_0=10km / h;

[0116] (3) Expressway mode M3:

[0117] Entry condition: Vehicle speed v>expressway vehicle speed threshold v_hw, v_hw=110km / h, and the duration reaches the set speed time threshold T_hold, and the day-night flag DayNight=night;

[0118] Exit condition: Vehicle speed v<expressway exit threshold v_tw, v_tw=100km / h;

[0119] (4) ADB adaptive high beam mode M4:

[0120] Entry condition: (Light switch state S_L=high beam or (light switch state S_L=AUTO, and day-night flag DayNight=night, and no oncoming target glare disable condition)), and camera validity flag Valid=1, and wiper gear G_w≤1, and gear signal R_g≠the set reverse gear R;

[0121] Exit condition: When the camera validity flag Valid=0, switch to normal high beam or low beam; when the wiper gear G_w≥2, switch to bad weather mode M6;

[0122] (5) Cornering Enhancement Mode M5 (works in conjunction with M1~M4):

[0123] Entry condition: Steering wheel angle Steering wheel angle threshold _act, _act=5°, and the duration reaches the set turning time threshold T_act, T_act=0.2s, and the vehicle speed v is within the curve speed range, the fusion curvature K>curvature threshold K_th, K_th=0.0051 / m;

[0124] Exit condition: Steering wheel angle <2° lasting for 1 second, or vehicle speed v is greater than the upper limit of the curve speed range;

[0125] (6) Severe Weather Mode M6:

[0126] Entry conditions: Wiper speed G_w ≥ 2, and the duration reaches the set wiper time threshold T_w, T_w = 30s;

[0127] Exit condition: Wiper setting G_w=0 for 60 seconds.

[0128] (7) Reversing auxiliary lighting mode M8:

[0129] Entry condition: Gear signal R_g = Reverse gear signal R; Exit condition: Gear signal R_g ≠ Reverse gear signal R.

[0130] (8) Welcome / Homecoming Mode M7: Triggered by unlock / lock events issued by the vehicle network, unrelated to perception data, and the duration is the calibration value, which is 15s in the example.

[0131] All the above "duration time T" conditions are implemented through a counter: if the condition expression is true, the counter value is incremented by 1 each control cycle; if it is false, it is reset to zero. The condition is determined to be true when the counter value reaches T / T_s. This mechanism, together with the hysteresis band, ensures the stability of mode switching and avoids light distribution jitter.

[0132] Table 3 shows the light pattern characteristics, Micro-LED pixel strategy, and AFS angle and brightness reference definitions for each mode.

[0133] Table 3. Benchmark Characteristics of Each Model

[0134]

[0135] The output states of the lighting mode management module and the control schemes corresponding to the calculation results of each state are summarized in Table 4, forming a complete mapping of "state calculation - control scheme". This mapping table is the logic for generating output instructions of the mode management module.

[0136] Table 4. State Calculation-Control Scheme Mapping Table

[0137]

[0138] In step 4b, the AFS angle control module calculates the horizontal deflection angle (curving lighting) and vertical deflection angle (slope compensation) of the headlights in real time based on sensing data such as steering wheel angle and vehicle speed, and generates motor execution commands to direct the beam of light along the road direction. The AFS angle control module consists of four sub-parts: a horizontal control unit, a vertical control unit, a two-level collaborative unit, and a motor closed-loop unit.

[0139] The horizontal control unit is used to perform dynamic curve lighting using a horizontal curve lighting algorithm, and the specific steps are as follows:

[0140] Step A1: The horizontal control unit first determines the steering wheel angle based on the steering wheel angle. The equivalent front wheel steering angle was calculated. The expression is:

[0141] ;

[0142] in, For the equivalent front wheel steering angle, This refers to the steering gear ratio;

[0143] Step A2: The horizontal control unit determines the equivalent front wheel steering angle based on the horizontal control unit. Given the vehicle speed v and yaw rate ω_z, the geometric curvature K_geo and yaw curvature K_yaw of the cornering lighting are calculated respectively, and their expressions are:

[0144] Geometric curvature: ;

[0145] Yaw curvature: This is only valid when the vehicle speed v > the minimum vehicle speed V_min, where V_min = 0.5 m / s;

[0146] Where L is the vehicle wheelbase. The understeer gradient is given by v, where v is the vehicle speed. This refers to the yaw rate;

[0147] Then, combining the navigation pre-aiming curvature K_nav, the fused curvature K is calculated, and its expression is:

[0148] ;

[0149] in, The signal validity indicator is the geometric curvature. This serves as a signal validity indicator for the yaw curvature. The value is 1 or 0; the denominator is the sum of the weights of each valid signal source; when the denominator is less than the set denominator threshold ε_κ, ε_κ=0.1, that is, all signal sources are ineffective or the confidence is too low, K=0 and the curve enhancement mode M5 is canceled to avoid division by zero and unfounded output;

[0150] The weights are assigned based on vehicle speed, as shown in Table 5:

[0151] Table 5

[0152]

[0153] At low speeds, the steering wheel angle has a high signal-to-noise ratio and weak yaw signal, so the geometric method is mainly used. At high speeds, navigation pre-aiming can know the curvature of the road ahead in advance and compensate for the lag in driver steering and vehicle posture response, so the weight is increased. When any signal fails, its weight is automatically reset to zero, achieving a seamless switching.

[0154] Step A3: The horizontal control unit determines the road aiming distance L_eff and the road curvature based on the road aiming distance L_eff. The target horizontal deflection angle was calculated. The expression is:

[0155] ;

[0156] AFS cornering pre-aiming geometry diagram is shown below. Figure 4 As shown, L_eff is the pre-aiming distance, and in this embodiment, L_eff = 25m; k_p is the pre-aiming geometric coefficient, the geometric derivation of which is as follows: Let the road curvature be κ, corresponding to the turning radius R = 1 / |κ|, the vehicle's longitudinal axis along the tangent direction of the arc, and the pre-aiming point P located at the arc length L_eff ahead. According to the tangent-chord theorem, the angle between the tangent at any point on the arc and the chord is equal to half the central angle subtended by the chord. The angle θ between the vehicle's longitudinal axis (the tangent) and the line connecting the vehicle to the pre-aiming point (the chord) satisfies θ = (L_eff•κ) / 2. Therefore, k_p = 0.5 is the analytical solution for the beam to accurately point to the pre-aiming point; when k_p takes a value in the interval (0.5, 1.0], the beam points to the curved area far from the pre-aiming point, which belongs to the aggressive lighting strategy of illuminating farther but reaching the limit earlier. k_p is calibrated according to the optical characteristics of the vehicle model and the lighting strategy requirements. In this embodiment, k_p = 0.5 is taken, which is the pre-aiming point criterion.

[0157] Then, the target horizontal deflection angle is... Perform adaptive speed limiting and gain correction to obtain the corrected target horizontal deflection angle. The expression is:

[0158] ;

[0159] in, For curvature gain, The target is the maximum horizontal deflection angle. A function that limits input values ​​to between a specified minimum and maximum value;

[0160] The curvature gain K_c(v) = max(0.6, 1 - 0.005•(v-40)), which decreases slightly with vehicle speed after v>40.

[0161] θ_h notation: a left turn is equivalent to the road curvature. <0 target angle If the value is negative, the lamp body deflects to the left, causing the beam to be directed towards the inside of the curve.

[0162] Step A4: The horizontal control unit adjusts the target horizontal deflection angle according to the corrected angle. Dead zone, rate limiting, and jitter reduction are performed.

[0163] When the absolute value of the corrected target horizontal deflection angle Target horizontal deflection angle threshold less than dead zone Update the current target horizontal deflection angle. Otherwise, update the current target horizontal deflection angle. ;

[0164] Then, based on the current target horizontal deflection angle... The horizontal mechanical target angle was calculated. The expression is:

[0165] ;

[0166] in, This is the maximum horizontal mechanical angular velocity.

[0167] That is, the output angle approaches the target angle with the maximum angular velocity R_h, taking into account both rapid response when entering a curve and smoothness under high-speed conditions. In the example, R_h = 20° / s when v < 40 km / h, and R_h = 8° / s when v ≥ 40 km / h.

[0168] The vertical control unit is used to perform slope compensation using a vertical slope compensation algorithm. The specific steps are as follows:

[0169] Step B1: The vertical control unit uses the front suspension height collected by the front and rear suspension height sensors. Rear suspension height Calculate the vehicle body pitch angle The expression is:

[0170] ;

[0171] in, Install longitudinal distance for the front and rear suspension height sensors. The front of the car should be tilted upwards;

[0172] When the pitch angle of the inertial measurement unit (IMU) When effective, take the vehicle body pitch angle. Otherwise, the vehicle's pitch angle ;

[0173] Step B2: Separation of static slope and dynamic pitch: The vertical control unit controls the vehicle's pitch angle. Perform a first-order low-pass filter to obtain the static component. The expression is:

[0174] ;

[0175] ;

[0176] in, For data frame indexing, These are the filter coefficients. For signal acquisition period, The filter time constant; static component Used to reflect the static pitch of road longitudinal slope and load.

[0177] Dynamic components The calculation expression is:

[0178] ;

[0179] Dynamic components Used to reflect the head-nodding and head-up transients caused by braking / acceleration.

[0180] And based on longitudinal acceleration The additional pitch compensation term was calculated. The expression is:

[0181] ;

[0182] in, The vehicle pitch component is estimated by longitudinal acceleration feedforward, with positive pitch during acceleration and negative pitch during braking. The notation conventions are consistent; To accelerate the pitch coupling coefficient; It represents longitudinal acceleration.

[0183] Step B3: The vertical control unit based on the static component Dynamic components And pitch compensation The target vertical deflection angle was calculated. The expression is:

[0184] ;

[0185] in, The dynamic component attenuation coefficient is K_d = 0.5 in this example, meaning that only partial compensation is made for transient pitch to avoid overshoot; For pitch-optical axis compensation ratio, The maximum target vertical deflection angle;

[0186] Compensation direction: The front of the vehicle tilts upward when going uphill. If the value is greater than 0, the optical axis will shift upwards. The value is negative, restoring the far-field illuminance to the level road reference; the same reverse compensation applies to downhill sections; the longitudinal acceleration feedforward term is output in advance for compensation during the initial stage of acceleration / braking transients, in conjunction with the dynamic component partial compensation coefficient. To suppress the glare caused by the upward tilt of the optical axis during acceleration and to avoid excessive downward tilt of the optical axis during braking and head-nodding, thus reducing the illumination distance.

[0187] Step B4: The vertical control unit determines the target vertical deflection angle based on the vertical deflection angle. The vertical mechanical target angle was calculated. The expression is:

[0188] ;

[0189] in, The maximum angular velocity in the vertical direction is used. Vertical compensation is a slow dynamic process, so a smaller angular velocity limit is used to ensure smoothness. The signal acquisition period; The vertical mechanical target angle.

[0190] The AFS mechanical deflection system drives the lamp body via a motor and a reduction mechanism, with a response time on the order of hundreds of milliseconds; the Micro-LED pixel-level brightness distribution switching requires only one refresh cycle of 5 ms. This invention cascades the two into a two-stage deflection system, using mechanical deflection to cover a large angle range for coarse adjustment, and using the overall translation of the pixel brightness distribution to cover small angle residuals for fine adjustment, eliminating mechanical lag.

[0191] The two-stage collaborative unit employs a two-stage beam deflection collaborative mechanism of "mechanical coarse adjustment + pixel fine adjustment" to perform an overall translation of the pixel brightness matrix output by the beam control module. The specific steps are as follows:

[0192] Step C1: The two-level collaborative unit provides feedback based on the position of the horizontal motor. and vertical motor position feedback The actual angle of the horizontal machinery was calculated. and vertical mechanical actual angle The expression is:

[0193] ;

[0194] ;

[0195] in, For the angular resolution of the horizontal position sensor, This refers to the angular resolution of the vertical position sensor.

[0196] Step C2: The two-level collaborative unit, based on the actual angle of the horizontal machinery... Vertical mechanical actual angle Current target horizontal deflection angle Vertical deflection angle to the target The horizontal deflection residuals were calculated separately. and vertical deflection residual The expression is:

[0197] Horizontal direction: ;

[0198] Vertical direction: ;

[0199] Step C3: The two-level cooperative unit adjusts the horizontal deflection residual accordingly. Vertical deflection residual The horizontal offset of each pixel was calculated. Pixel vertical offset The expression is:

[0200] level: ;

[0201] vertical: ;

[0202] And limit the amplitude n_shift ≤ N_max; ;

[0203] in, This represents the horizontal angular resolution corresponding to a single column of the pixel matrix. N_max represents the vertical angular resolution corresponding to a single column of the pixel matrix, and N_max represents the maximum number of columns for pixel offset. This represents the maximum number of rows for pixel offset.

[0204] Step C4: The two-level collaborative unit adjusts the pixel horizontal offset accordingly. Vertical offset of pixels The overall translation amount of the pixel brightness matrix is ​​calculated. The expression is:

[0205] ;

[0206] , When the value exceeds the matrix range, the corresponding column is set to 0.

[0207] The translation is completed within a 5ms pixel refresh cycle, so the beam direction is in place before the mechanical motor reaches the target angle, and then the mechanical angle gradually approaches the target. convergence to Then it resets to zero, and the system returns to a purely mechanical deflection state. The two-stage beam deflection coordination mechanism of "mechanical coarse adjustment + pixel fine adjustment" reduces the beam pointing response time of curved illumination from about 200ms mechanical response time to the order of 5ms per pixel refresh cycle. At the same time, the mechanical angle undertakes the main deflection range, avoiding pixel translation that would sacrifice the illumination field of view.

[0208] The motor closed-loop unit is used for motor closed-loop control and execution output. The specific steps are as follows:

[0209] Step D1: The motor closed-loop unit determines the horizontal mechanical target angle based on the target angle. and vertical mechanical target angle Generate the corresponding target position for the motor, expressed as:

[0210] level: ;

[0211] vertical: ;

[0212] in, The target horizontal position for the motor. The motor is positioned perpendicular to the target. For the angular resolution of the horizontal position sensor, This refers to the angular resolution of the vertical position sensor.

[0213] Step D2: The motor closed-loop unit calculates the horizontal position error for each motor control cycle. and vertical position error The expression is:

[0214] ;

[0215] ;

[0216] in, The target horizontal position for the motor. For horizontal motor position feedback, The motor is positioned perpendicular to the target. For vertical motor position feedback;

[0217] Then based on the horizontal position error and vertical position error The horizontal control quantity u and the vertical control quantity u_v of the output motor speed duty cycle are expressed as follows:

[0218] ;

[0219] ;

[0220] Positional error is expressed in mechanical degrees. , For the horizontal position loop gain, , For vertical position loop gain, For motor control cycle, For normalized speed commands, ; This represents the horizontal position error of the previous motor control cycle. The vertical position error of the previous motor control cycle is output to the AFS motor driver via PWM, and its first output group OUT1A / OUT1B drives the left headlight adjustment motor, and the second output group OUT2A / OUT2B drives the right headlight adjustment motor.

[0221] Step D3: When configuring different deflection characteristics for low beam and high beam headlights according to target market regulations such as ECE R123 for Class C system cornering lighting deflection characteristics and GB 21259 corresponding limits, the target angles for the left and right lights can be calculated separately: the inner light (closest to the bend) takes... The outer light is 0.6. To achieve asymmetrical beam pattern deployment in curves;

[0222] Step D4: When the absolute value of the position error When the set position error threshold e_th is set, and the motor current I_mot is greater than the set motor current threshold I_th, and the duration reaches the stall time threshold t_stall, the motor is determined to be stalled, a stall fault flag is set, and the safety management module is notified to perform a downgrade.

[0223] In step 4c, the Micro-LED beam control module takes the camera target list and the AFS mechanical actual angle as input, outputs the brightness setting value of each pixel in the Micro-LED pixel matrix, and executes it through the driving circuit. The specific steps are as follows:

[0224] Step E1: Target Selection: The Micro-LED beam control module identifies each target in the camera's target list and retains targets that meet the following conditions for inclusion in the occlusion calculation:

[0225] T_j∈{oncoming vehicles, same-direction vehicles, pedestrians}, and c_j>c_th, and 10≤d_j≤250m, and | |Located within the field of view of the forward-facing camera;

[0226] Where T_j represents the type of the j-th target, such as oncoming vehicle, same-direction vehicle, pedestrian, or others; c_j represents the recognition confidence of the j-th target, c_j∈(0,1); c_th represents the confidence threshold; and d_j represents the distance to the j-th target.

[0227] Step E2a: Target Motion Feedforward Prediction: To compensate for the changes in target angle and distance during the total delay T_lat introduced by camera image processing, target recognition, and bus transmission, the Micro-LED beam control module calculates the azimuth angle change rate for each effective target. Pitch angle change rate and distance change rate The expression is:

[0228] ;

[0229]

[0230] ;

[0231] Where T_cam is the camera frame period and N_pred is the differential frame interval; The rate of change of azimuth angle. The rate of change of distance, Here, j is the data frame index, and j is the camera target index; The azimuth angle of the target relative to the longitudinal axis of this vehicle is set to right deviation as positive; The pitch angle of the target relative to the vehicle's horizontal baseline is set to positive for upward deviation. Distance to target;

[0232] Then, based on the azimuth rate of change... Pitch angle change rate and distance change rate The predicted azimuth angle is calculated. Pitch angle prediction value and distance prediction values The expression is:

[0233] ;

[0234] ;

[0235] ;

[0236] in, Total latency introduced for camera image processing, target recognition, and bus transmission;

[0237] Step E2b: Lamp coordinate transformation for target angle: The target angle output by the camera is based on the vehicle coordinate system, while the pixel matrix rotates mechanically with the lamp body. The Micro-LED beam control module converts the predicted azimuth angle value... Pitch angle prediction value Transform to the lamp body coordinate system; the expression is:

[0238] ;

[0239] ;

[0240] in, The actual angle of the horizontal machinery. For the actual angle of the vertical machinery The target azimuth angle in the lamp body coordinate system. The target elevation angle in the lamp coordinate system;

[0241] The mechanical actual angle conversion ensures that the shielded area remains locked in the road coordinate system for traffic participants during mechanical rotation, thus solving the problem of shielded area drift caused by mechanical deflection.

[0242] Step E3: Target angle to pixel coordinate mapping: The Micro-LED beam control module calculates the target center pixel coordinates using the center of the pixel matrix as the optical axis reference. The expression is:

[0243] ;

[0244] ;

[0245] in, The number of the center column of the matrix. The center row number of the matrix. This represents the horizontal angular resolution corresponding to a single column of the pixel matrix. This represents the vertical angular resolution corresponding to a single row of the pixel matrix. =0.05° / column, =0.15° / row This is the rounding function. The x-coordinate of the center pixel of the target The value is the ordinate of the center pixel of the target; the calculation result is rounded and limited to the range of the matrix.

[0246] Step E4: Distance-Adaptive Masking Template Generation: The Micro-LED beam control module generates a rectangular masking template M_j for each valid target, the horizontal half-width of which is... The calculation expression is:

[0247] ;

[0248] in, The horizontal occlusion width of the target. For angle safety margin;

[0249] The template's vertical range extends upwards from the row containing the target image to above the high beam cutoff line. The expression is:

[0250] ;

[0251] ;

[0252] ;

[0253] in, This indicates the lower boundary signal of the masking template. This represents the y-coordinate of the center pixel of the j-th target. This indicates the signal at the upper boundary of the masking template. This indicates the number of rows at half the vertical height of the masking template. This indicates a safety margin for vertical upward movement. This represents the vertical angular resolution corresponding to a single row of the pixel matrix;

[0254] When |c-col_j|≤n_w,j and row_top≤r≤row_bottom, the single-target occlusion template M_i(r,c)=0, and the template is turned off; otherwise, the single-target occlusion template M_j(r,c)=1, and the template is kept. Here, c represents the pixel column index, col_j represents the horizontal coordinate of the center pixel of the j-th target, r represents the pixel row index, and n_w,j represents the horizontal half-width of the occlusion template of the j-th target.

[0255] Steps E3-E4 calculate pixel coordinates and occlusion template size using the predicted target angle and distance. This feedforward ensures that the occlusion area reaches the pixel region the target is about to enter in advance during high-speed encounters between two vehicles, eliminating the phenomenon that occlusion establishment lags behind target movement; (dd / dt)_j also serves as the input for the opposing / same-direction discrimination in step E10, and the equivalent prediction time can be further extended for opposing targets to enable earlier occlusion. The calculation results are rounded and limited to the matrix range (col∈[1,320],row∈[1,80]).

[0256] Step E5: Multi-target template synthesis: The Micro-LED beam control module performs a logical AND operation on a pixel-by-pixel basis on the masking templates of all valid targets to synthesize the total masking template.

[0257] M(r,c)=min(M_1(1,c),M_2(r,c),...,M_n(r,c));

[0258] Where M(r,c) represents the total occlusion template, M_n(r,c) is the occlusion template of the nth valid target, and n is the total number of valid targets;

[0259] Step E6: Temporal anti-flicker processing: Asymmetric timing is used for occlusion establishment and de-occlusion.

[0260] Setup: When the frame takes effect -- within the same refresh cycle when the target appears, immediately output the total occlusion template M(r,c)=0, with anti-glare priority and a response time of 5ms;

[0261] Unmasking: After the target disappears, the occlusion template remains for 300ms, and then gradually fades linearly within the unmasking time threshold T_fade=200ms.

[0262] M_out(r,c,k)=min(M(r,c,k),M_hold(r,c,k-1));

[0263] Where M_hold(r,c,k-1) means to maintain the occlusion state of the previous frame within the window, M(r,c,k) is the occlusion state of the current frame, and M_out(r,c,k) is the total occlusion template output of the current frame;

[0264] This asymmetric timing prevents detection flicker, i.e., the flickering of lights and visual discomfort caused by the high-frequency switching of the shielded area due to the brief missed detection of the target.

[0265] Step E7: Softening the edges of the occluded area: using a Gaussian convolution kernel Output of total masking template The edges are softened to prevent hard edges from projecting onto the road surface and forming abrupt dark patches. This softened overall masking template output is then obtained. The expression is:

[0266] ;

[0267] in, for Gaussian convolution kernel of pixels, This is a convolution operation, which only operates within a 1-2 pixel width of the occlusion boundary;

[0268] Step E8: Pixel Luminance Combination: Final Pixel Luminance Matrix The calculation expression is:

[0269] ;

[0270] Wherein, L_base is the mode reference luminance matrix, which is output by the mode management module; M_soft is the temperature derating factor, and M_soft is the anti-glare shielding factor, where M_soft∈(0,1); This represents the overall pixel translation amount;

[0271] Step E9: Driving Parameter Mapping and Distribution: Mapping the final pixel brightness matrix Converted into drive circuit control quantities, and pixel-level drive control is adopted;

[0272] Step E10: Oncoming / Same-Direction Target Determination: When the target distance change rate dd_j / dt < the oncoming vehicle change rate threshold -8 m / s, the target is determined to be an oncoming vehicle; when the target distance change rate dd_j / dt is within the same-direction vehicle change rate range [-3, +8] m / s and is located in front of the vehicle in the same lane, the target is determined to be a same-direction vehicle; a full masking template is applied to oncoming vehicles; for same-direction vehicles, only the area above the reflector height is masked (row_top is moved up 1 pixel); pedestrians are fully masked with increased vertical margin. A schematic diagram of the masking template is shown below. Figure 3 As shown.

[0273] The Micro-LED beam control module takes the camera target list and the actual AFS mechanical angle as input, and outputs the brightness setting value of each pixel in the Micro-LED pixel matrix, which is then executed by the driver circuit. The pixel matrix has a size of 320 columns × 80 rows, 25,600 pixels, and uses an ams OSRAM EVIYOS 2.0 light source. The horizontal angular resolution of a single pixel is Δα_px = 0.05° / pixel, and the vertical angular resolution is Δβ_px = 0.15° / pixel. The beam pattern covers ±8° horizontally and 12° vertically, and the independent pixel setting is achieved through the driver.

[0274] In step 4d, the safety management module is used to continuously monitor the system operating status and perform temperature derating, power budget allocation, and graded failure degradation control, as follows:

[0275] (1) Channel-level monitoring and protection:

[0276] Overcurrent / short circuit protection: When the current I_ch of any channel is greater than 1.5 times the protection circuit threshold I_rated and the duration reaches the set protection current time threshold of 5ms, the channel will be immediately shut down and the channel fault flag will be set; when 3 or more channels fail consecutively, the system will enter L5 level degradation.

[0277] Open circuit / pixel failure detection: Use current sampling during row gating to detect open circuits column by column, mark columns with channel current below the threshold as open-circuit pixels, and replace open-circuit pixels with adjacent column interpolation in the brightness matrix to avoid dark spots;

[0278] (2) Temperature derating and power budget allocation:

[0279] Temperature derating: Calculate the global brightness reduction coefficient according to the lamp body temperature T_sen , the expression is:

[0280] When the lamp body temperature derating temperature threshold , , ;

[0281] otherwise, ;

[0282] That is, after the temperature exceeds 80°C, the global brightness decreases by 2% for every 1°C increase, and the minimum drops to 50%; the global brightness reduction coefficient is input into the Micro-LED beam control module to participate in brightness synthesis; when the temperature exceeds 105°C, the cooling fan is started at full speed and further derating is performed;

[0283] Power budget allocation: Estimate the total required power of the system in real time , the expression is:

[0284] ;

[0285] where, is the AFS motor power, is the high-side load power such as fans, heaters, etc., is the channel voltage, is the channel current, is the PWM duty cycle of the LED channel;

[0286] When the total required power > the set total power threshold , reduce the adjustable loads in order from the lowest to the highest function priority;

[0287] The priority order is: decorative / welcome light pattern < projection display < ADB far-field pixel < curve supplementary light pixel < low beam basic pixel;

[0288] The calculation expression of the power attenuation coefficient is:

[0289] =1-( ) / ;

[0290] in, It provides adjustable power; it only applies to pixel groups with a lower priority than the currently protected object, ensuring that the basic low beam distribution is not affected by power management.

[0291] Degradation state machine transition conditions: Degradation level only increases and does not jump to recovery -- when the fault severity is higher than the current level, it immediately increases to the corresponding level; during recovery, it only drops back one level each time, and the backflow condition is that the fault disappears and the duration reaches the recovery time threshold T_recover, T_recover=10s.

[0292] The security management module implements tiered failure degradation control, specifically including the following degradation control schemes:

[0293] Level LO: Normal; Triggering condition: No fault; Degradation control scheme: All system functions are executed normally;

[0294] Level L2: Camera downgrade; Triggering condition: Camera target list timeout, or camera validity flag Valid=0; Downgrade control scheme: Exit ADB mode M4, switch to low beam M1, AFS cornering lighting and slope compensation are maintained;

[0295] Level L3: Mechanical execution downgrade; Triggering condition: Motor stall judgment, or position feedback failure; Downgrade control scheme: Disconnect the motor drive output to make the lamp body return to the center via the reset spring, pixel offset limit, N_max is increased from 10 columns to 40 columns, and the curve beam deflection is handled by the Micro-LED pixel translation;

[0296] Level L4: Communication degradation; Triggering conditions: CAN error count on the environmental bus > the set error count threshold Bus0ff, or the task bus fails; Degradation control scheme: After maintaining the last effective mode for 3 seconds, switch to the default low beam (safe state), and the local temperature protection and low beam constant current drive continue to work.

[0297] Level L5: Critical fault; Trigger condition: Overheating, i.e., lamp body temperature. >115°C, or a constant current fault or power supply abnormality in 3 or more channels; Degradation control scheme: only retain the constant current channel corresponding to the basic low beam, reduce the brightness to 50%, report the diagnostic fault code (DTC) and request the whole vehicle to switch to emergency lighting.

[0298] Each level of degradation judgment condition includes a duration requirement to filter out transient interference; both degradation events and recovery events generate diagnostic codes (DTCs) and are uploaded to the vehicle network via the vehicle environment bus.

[0299] The system defines six degradation states. When a fault occurs, the system enters the corresponding level according to its severity, tightening functions step by step. After the fault is recovered, the system gradually revs up after T_recover = 10 s confirmation. The correspondence between state levels, triggering conditions, and degradation control schemes is shown in Table 6.

[0300] Table 6. Correspondence between Status Levels, Triggering Conditions, and Degradation Control Schemes

[0301]

[0302] The execution module includes an LED driving circuit, a high-side driving circuit, and a motor driving circuit. The MCU main controller drives the MicroLED combination via the LED driving circuit, drives the high-side load via the high-side driving circuit, and drives the AFS execution motor via the motor driving circuit.

[0303] The following are the verification data of each algorithm module in this embodiment under the simulation environment:

[0304] (1) Multi-source curvature fusion: Under the constructed curve condition, the geometric source contains a 0.4 s steering establishment lag and a 12% steady-state deviation, the yaw source contains strong noise, and the navigation source contains quantification error and a confidence level of 0.8. The steady-state curvature RMS error of the fusion estimate is less than 68% of that of the optimal single source, and the accuracy is improved by more than 32%. When a single geometric source fails midway, the mean deviation of the fusion output before and after switching is less than 1× 1 / m, no jump.

[0305] (2) Two-level coordination: Under the step condition of 5° target angle, the pixel matrix completes 0.5°, i.e. 10 columns, in the first 5 ms refresh cycle after the step. The mechanical actuator then follows up and arrives in place in about 200 ms. After arriving in place, the deflection residual converges to within 0.05° and the pixel translation automatically returns to zero. Under the L3 degradation condition of motor failure, the pure pixel electronic compensation reaches 2.0°, i.e. 40 columns.

[0306] (3) Pixel-level anti-glare: When an oncoming vehicle approaches at a relative speed of 25 m / s and the total image link delay is 100 ms, the target motion feedforward prediction reduces the pixel mapping error RMS from about 1.9 columns to about 1.1 columns, and the occlusion coverage of the pixel column where the target is located remains 100%; when the lamp body is mechanically deflected by 1°, the target pixel coordinates are synchronously corrected by about 20 columns, verifying the locking and following of the occlusion area in the road coordinate system.

[0307] (4) Vertical slope compensation: The steady-state compensation error for a 2° static uphill slope is less than 0.3°; acceleration +2 m / With braking -3 m / In the transient state, the longitudinal acceleration feedforward outputs the correct direction compensation within 0.5 s, and the output is always within the ±3° limit.

[0308] (5) Motor closed-loop and stall detection: the target tracking error of the slope is less than 0.5°; after the mechanism is stuck, the stall flag is triggered within the 0.5 s criterion margin, and the large error at the beginning of the slope with short duration will not be falsely reported.

[0309] (6) Mode management and safety management: speed peaks that are not held for long enough will not trigger high-speed mode; day and night indicators will not shake when ambient light illuminance fluctuates within the hysteresis band of 500-700 lx; ​​total power will not exceed budget after power reduction and low beam basic power will be fully retained; the six-level downgrade state machine will be upgraded according to events and will gradually fall back after fault recovery with 10s confirmation.

[0310] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for cooperative control of a Micro-led digital headlamp and an AFS actuator, characterized in that, Includes the following steps: Step 1: Construct a collaborative control system for Micro-led digital headlights and AFS actuators. This collaborative control system is equipped with an MCU main controller, which is connected to a data acquisition module and an execution module. The MCU main controller integrates a preprocessing module, a lighting mode management module, an AFS angle control module, a Micro-LED beam control module, and a safety management module; Step 2: The data acquisition module periodically acquires vehicle data vector x and transmits it to the MCU main controller; Step 3: The preprocessing module in the MCU main controller performs preprocessing operations on the vehicle data vector x to obtain the filtered effective value and validity flag of each data. Step 4a: The lighting mode management module determines the lighting mode based on the effective value and validity flag of each data, obtains the current vehicle lighting mode, and then outputs the corresponding lighting control scheme to the execution module according to the lighting mode. Step 4b: The AFS angle control module calculates the headlight horizontal and vertical deflection angles in real time according to the steering wheel angle and vehicle speed v in the vehicle data vector x, and generates motor execution instructions to the execution module. and vehicle speed v, and generates motor execution instructions to the execution module. Step 4c: The Micro-LED beam control module outputs a Micro-LED light driving command to the execution module based on the camera target list in the vehicle data vector x and the AFS mechanical actual angle; Step 4d: The security management module generates a security level upgrade / downgrade control command for the execution module based on the filtered effective value and validity flag of each data. Step 5: The execution module implements coordinated control and drive of the Micro-led digital headlights and AFS actuators based on the lighting control scheme, motor execution instructions, Micro-LED lighting drive instructions, and safety level adjustment control instructions.

2. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 1, characterized in that: In the step 2, the vehicle data vector x includes but is not limited to the steering wheel angle of the vehicle , the vehicle speed v, the yaw rate ω_z, the longitudinal acceleration , the wiper gear G_w, the light switch state S_L, the gear signal R_g, the road type RoadType, the navigation preview curvature K_nav, the camera effective target list, the motor position feedback P_fb data, the ambient light intensity , the lamp body temperature T_sen, the channel current I_ch, the channel voltage U_ch data, the front suspension height h_f and the rear suspension height h_r; The data acquisition module is equipped with a vehicle environment bus, a task execution bus, and an ADC sampling circuit. The MCU main controller is connected with a vehicle body controller and an intelligent driving domain controller through a vehicle body environment bus, and is used for acquiring a steering wheel rotation angle , a vehicle speed v, a yaw angular velocity ω_z, a longitudinal acceleration , a wiper gear G_w, a light switch state S_L, a gear signal R_g, a road type RoadType, a navigation preview curvature K_nav and camera effective target list data; The MCU main controller is connected to the AFS actuator via a task execution bus to acquire the vehicle's motor position feedback P_fb data; The MCU main controller acquires ambient light intensity through an ADC sampling circuit , lamp body temperature T_sen, channel current I_ch and channel voltage U_ch data.

3. The Micro-led digital headlamp and AFS actuator cooperative control method according to claim 1, characterized in that: The preprocessing module is provided with a limiting filter unit, a low-pass filter unit, and a validity flag and timeout monitoring unit connected in sequence. A dead zone and normalization processing unit is also provided between the low-pass filter unit and the validity flag and timeout monitoring unit. In the step 3, the preprocessing module performs preprocessing operation on each original signal in the vehicle data vector x, the specific steps are as follows: performing preprocessing operation on each original signal in the vehicle data vector x, the specific steps are as follows: Step S1: the amplitude limiting filter unit performs amplitude limiting filtering on each original signal collected per cycle If the difference between the current frame signal and the last valid frame signal is greater than the maximum change rate limit value of a single cycle of the signal , the current frame signal is determined as an abnormal frame signal and discarded, and the last valid frame signal is output as the current valid frame signal ; otherwise, the current frame signal is output as the current valid frame signal , and the logic expression is: When then ; Otherwise, ; Where k represents the sampling frame index and i represents the original signal type index; Step S2: the low-pass filter unit carries out first-order inertia low-pass filtering on the current effective frame signal passing through the amplitude limiting filter unit to obtain the filtered effective value of each signal , and the expression is: ; ; wherein, is the filtered effective value of the kth frame of the ith original signal, is the signal acquisition period, is the filter time constant; Step S3: The dead zone and normalization processing unit filters the filtered effective value of the steering wheel angle in the original signal The dead zone processing is performed, and the expression is when hour, ; otherwise, ; in, This is the effective value of the filtered steering wheel angle. The steering wheel angle after dead zone processing. The dead zone steering wheel angle threshold. It is a symbolic function; The dead zone and normalization processing unit performs logarithmic normalization on the effective value E of the ambient light intensity in the original signal, as expressed by: ; in, The normalized ambient illuminance This represents the effective value of the filtered ambient light intensity. This is the normalized reference illuminance parameter; Step S4: The validity flag and timeout monitoring unit monitor each signal in real time. When a monitored signal meets any of the following conditions, the validity flag of that signal is set to invalid, Valid=0: (1) The number of consecutive dropped frames of the signal is greater than or equal to the set frame number threshold. ; (2) Communication timeout: No update for more than 3 data collection cycles; (3) Exceeding the physical range: The signal value exceeds the range specified in the table above; (4) Verification error: CAN message CRC verification failed or ADC self-test abnormal sensor fault flag; Otherwise, the signal is marked as valid, Valid=1; Step S5: The preprocessing module finally outputs the filtered effective values ​​of each signal and their effectiveness flags, as well as the steering wheel angle after dead zone processing. Normalized ambient illuminance .

4. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 1, characterized in that: In said step 4a, said light modes include 8 light modes, which are respectively: off / parking light mode M0, low beam basic mode M1, urban mode M2, highway mode M3, ADB adaptive high beam mode M4, curve enhancement mode M5, bad weather mode M6, welcome / escort home mode M7, and reverse auxiliary lighting mode M8; The output light mode at any moment is determined by the light mode that satisfies the entry condition and has the highest priority, and the priorities of the light modes are as follows: P(M8)>P(M7)>P(M6)>P(M4)>P(M3)>P(M2)>P(M1)>P(M0); The curve enhancement mode M5 is a parallel superimposed mode, which does not participate in the above priority arbitration, but is superimposed as a modification layer on the current ground state modes M1 to M4; The input data for the lighting mode management module includes: vehicle speed v, ambient light intensity E, light switch status S_L, wiper gear position G_w, gear position signal R_g, road type RoadType, number of valid targets from the camera N_obj, and steering wheel angle. The navigation aiming curvature K_nav, and the system power-on / power-off status. The light mode determination logic of said light mode management module is as follows: (1) Automatic light activation, the prerequisite for the entry of the low beam basic mode M1 is that the light switch state is in the automatic state, that is, when the light switch state S_L=AUTO: When E<E_on=500Lx, set the day-night flag DayNight=night; When E>E_off=700Lx and the duration reaches the set illumination time threshold T_E, set the day-night flag DayNight=day; Wherein, E represents the ambient illuminance, E_on represents the automatic light activation enable threshold, and E_off represents the automatic light activation disable threshold; (2) Urban mode M2: Entry conditions: vehicle speed v<vehicle speed threshold v_city, the duration reaches the set speed time threshold T_hold, the road type RoadType is urban area or the number of valid camera targets N_obj≥the set target number threshold, and the day-night flag DayNight=night; Exit conditions: vehicle speed v>vehicle speed threshold v_city+hysteresis vehicle speed v_0 or the road type RoadType changes to highway; (3) Highway mode M3: Entry conditions: vehicle speed v>highway vehicle speed threshold v_hw, the duration reaches the set speed time threshold T_hold, and the day-night flag DayNight=night; Exit condition: vehicle speed v<highway exit threshold v_tw; (4) ADB adaptive high beam mode M4: Entry conditions: (the light switch state S_L=high beam or (the light switch state S_L=AUTO, the day-night flag DayNight=night, and no oncoming target glare disable condition)), the camera validity flag Valid=1, the wiper gear G_w≤1, and the gear position signal R_g≠the set reverse gear R; Exit conditions: when the camera validity flag Valid=0, switch to normal high beam or low beam; when the wiper gear G_w≥2, switch to the bad weather mode M6; (5) Curve enhancement mode M5: Entry condition: Steering wheel angle Steering wheel angle threshold _act, and the duration reaches the set corner time threshold T_act, and the vehicle speed v is within the curve speed range, and the fusion curvature K>curvature threshold K_th; Exit condition: Steering wheel angle <2° lasting for 1 second, or vehicle speed v is greater than the upper limit of the curve speed range; (6) Bad weather mode M6: Entry conditions: the wiper gear G_w≥2, and the duration reaches the set wiper time threshold T_w; Exit condition: the wiper gear G_w=0, and lasts for 60s. (7) Reverse auxiliary lighting mode M8: Entry condition: Gear signal R_g = Reverse gear signal R; Exit condition: Gear signal R_g ≠ Reverse gear signal R. (8) Welcome / Homecoming Mode M7: Triggered by unlock / lock events issued by the vehicle network, unrelated to perception data, and the duration is the calibrated value.

5. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 1, characterized in that: In step 4b, the AFS angle control module is equipped with a horizontal control unit, a vertical control unit, a two-level collaborative unit, and a motor closed-loop unit; The horizontal control unit is used to perform dynamic curve lighting using a horizontal curve lighting algorithm, and the specific steps are as follows: Step A1: The horizontal control unit first determines the steering wheel angle based on the steering wheel angle. The equivalent front wheel steering angle was calculated. The expression is: ; in, For the equivalent front wheel steering angle, This refers to the steering gear ratio; Step A2: The horizontal control unit determines the equivalent front wheel steering angle based on the horizontal control unit. Given the vehicle speed v and yaw rate ω_z, the geometric curvature K_geo and yaw curvature K_yaw of the cornering lighting are calculated respectively, and their expressions are: Geometric curvature: ; Yaw curvature: ; Where L is the vehicle wheelbase. The understeer gradient is given by v, where v is the vehicle speed. This refers to the yaw rate; Then, combining the navigation pre-aiming curvature K_nav, the fused curvature K is calculated, and its expression is: ; in, The signal validity indicator is the geometric curvature. This serves as a signal validity indicator for the yaw curvature. The denominator is either 1 or 0; the sum of the weights of each effective signal source is the denominator. Step A3: The horizontal control unit determines the road aiming distance L_eff and the road curvature based on the road aiming distance L_eff. The target horizontal deflection angle was calculated. The expression is: ; Where L_eff is the road preview distance, and k_p is the preview geometry coefficient. For road curvature; Then, the target horizontal deflection angle is... Perform adaptive speed limiting and gain correction to obtain the corrected target horizontal deflection angle. The expression is: ; in, For curvature gain, The target is the maximum horizontal deflection angle. A function that limits input values ​​to between a specified minimum and maximum value; Step A4: The horizontal control unit adjusts the target horizontal deflection angle according to the corrected angle. Dead zone, rate limiting, and jitter reduction are performed. When the absolute value of the corrected target horizontal deflection angle Target horizontal deflection angle threshold less than dead zone Update the current target horizontal deflection angle. Otherwise, update the current target horizontal deflection angle. ; Then, based on the current target horizontal deflection angle... The horizontal mechanical target angle was calculated. The expression is: ; in, The maximum horizontal mechanical angular velocity; The vertical control unit is used to perform slope compensation using a vertical slope compensation algorithm. The specific steps are as follows: Step B1: The vertical control unit adjusts according to the front suspension height. Rear suspension height Calculate the vehicle body pitch angle The expression is: ; in, Install longitudinal distance for the front and rear suspension height sensors. The front of the car should be tilted upwards; When the pitch angle of the inertial measurement unit (IMU) When effective, take the vehicle body pitch angle. Otherwise, the vehicle's pitch angle ; Step B2: Separation of static slope and dynamic pitch: The vertical control unit controls the vehicle's pitch angle. Perform a first-order low-pass filter to obtain the static component. The expression is: ; ; in, For data frame indexing, These are the filter coefficients. For signal acquisition period, The filtering time constant; Dynamic components The calculation expression is: ; And based on longitudinal acceleration The additional pitch compensation term was calculated. The expression is: ; in, The vehicle pitch component is estimated by longitudinal acceleration feedforward. To accelerate the pitch coupling coefficient; Indicates longitudinal acceleration; Step B3: The vertical control unit based on the static component Dynamic components And pitch compensation The target vertical deflection angle was calculated. The expression is: ; in, The dynamic component attenuation coefficient, For pitch-optical axis compensation ratio, The maximum target vertical deflection angle; Compensation direction: The front of the vehicle tilts upward when going uphill. If the value is greater than 0, the optical axis will shift upwards. The value is negative, restoring the far-field illuminance to the level road reference; the same reverse compensation applies to downhill sections; the longitudinal acceleration feedforward term is output in advance for compensation during the initial stage of acceleration / braking transients, in conjunction with the dynamic component partial compensation coefficient. To suppress the glare caused by the upward tilt of the optical axis during acceleration and to avoid excessive downward tilt of the optical axis during braking and head-nodding, thus reducing the illumination distance. Step B4: The vertical control unit determines the target vertical deflection angle based on the vertical deflection angle. The vertical mechanical target angle was calculated. The expression is: ; in, The maximum angular velocity in the vertical direction. The signal acquisition period; The vertical mechanical target angle; The two-stage collaborative unit employs a two-stage beam deflection collaborative mechanism of "mechanical coarse adjustment + pixel fine adjustment" to perform an overall translation of the pixel brightness matrix output by the beam control module. The specific steps are as follows: Step C1: The two-level collaborative unit provides feedback based on the position of the horizontal motor. and vertical motor position feedback The actual angle of the horizontal machinery was calculated. and vertical mechanical actual angle The expression is: ; ; in, For the angular resolution of the horizontal position sensor, This refers to the angular resolution of the vertical position sensor. Step C2: The two-level collaborative unit, based on the actual angle of the horizontal machinery... Vertical mechanical actual angle Current target horizontal deflection angle Vertical deflection angle to the target The horizontal deflection residuals were calculated separately. and vertical deflection residual The expression is: Horizontal direction: ; Vertical direction: ; Step C3: The two-level cooperative unit adjusts the horizontal deflection residual accordingly. Vertical deflection residual The horizontal offset of each pixel was calculated. Pixel vertical offset The expression is: level: ; vertical: ; And limit the amplitude n_shift ≤ N_max; ; in, This represents the horizontal angular resolution corresponding to a single column of the pixel matrix. N_max represents the vertical angular resolution corresponding to a single column of the pixel matrix, and N_max represents the maximum number of columns for pixel offset. This represents the maximum number of rows for pixel offset. Step C4: The two-level collaborative unit adjusts the pixel horizontal offset accordingly. Vertical offset of pixels The overall translation amount of the pixel brightness matrix is ​​calculated. The expression is: ; The motor closed-loop unit is used for motor closed-loop control and execution output. The specific steps are as follows: Step D1: The motor closed-loop unit determines the horizontal mechanical target angle based on the target angle. and vertical mechanical target angle Generate the corresponding target position for the motor, expressed as: level: ; vertical: ; in, The target horizontal position for the motor. The motor is positioned perpendicular to the target. For the angular resolution of the horizontal position sensor, This refers to the angular resolution of the vertical position sensor. Step D2: The motor closed-loop unit calculates the horizontal position error for each motor control cycle. and vertical position error The expression is: ; ; in, The target horizontal position for the motor. For horizontal motor position feedback, The motor is positioned perpendicular to the target. For vertical motor position feedback; Then based on the horizontal position error and vertical position error The horizontal control quantity u and the vertical control quantity u_v of the output motor speed duty cycle are expressed as follows: ; ; in, , For the horizontal position loop gain, , For vertical position loop gain, For motor control cycle, For normalized speed commands, ; This represents the horizontal position error of the previous motor control cycle. This represents the vertical position error of the previous motor control cycle; Step D3: When the low beam and high beam are configured with different deflection characteristics according to the regulations of the target market, the target angle of the left and right lights is calculated separately: the inner light is taken as... The outer light is 0.

6. To achieve asymmetrical beam pattern deployment in curves; Step D4: When the absolute value of the position error When the set position error threshold e_th is set, and the motor current I_mot is greater than the set motor current threshold I_th, and the duration reaches the stall time threshold t_stall, the motor is determined to be stalled, a stall fault flag is set, and the safety management module is notified to perform a downgrade.

6. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 1, characterized in that: In step 4c, the Micro-LED beam control module takes the camera target list and the AFS mechanical actual angle as input and outputs the brightness setting value of each pixel in the Micro-LED pixel matrix. The specific steps are as follows: Step E1: Target Selection: The Micro-LED beam control module identifies each target in the camera's target list and retains targets that meet the following conditions for inclusion in the occlusion calculation: T_j∈{oncoming vehicles, same-direction vehicles, pedestrians}, and c_j>c_th, and 10≤d_j≤250m, and | |Located within the field of view of the forward-facing camera; Where T_j represents the type of the j-th target, c_j represents the confidence level of the j-th target, c_th represents the confidence threshold, and d_j represents the distance to the j-th target; Step E2a: Target motion feedforward prediction: The Micro-LED beam control module calculates the azimuth rate of change for each effective target. Pitch angle change rate and distance change rate The expression is: ; ; Where T_cam is the camera frame period and N_pred is the differential frame interval; The rate of change of azimuth. The rate of change of distance, Here, j is the data frame index, and j is the camera target index; The target azimuth angle, For the target pitch angle, The target distance; Then, based on the azimuth rate of change... Pitch angle change rate and distance change rate The predicted azimuth angle is calculated. Pitch angle prediction value and distance prediction values The expression is: ; ; ; in, Total latency introduced for camera image processing, target recognition, and bus transmission; Step E2b: Lamp coordinate transformation for target angle: The Micro-LED beam control module will convert the predicted azimuth angle value... Pitch angle prediction value Transform to the lamp body coordinate system; the expression is: ; ; in, The actual angle of the horizontal machinery. For the actual angle of the vertical machinery The target azimuth angle in the lamp body coordinate system. The target elevation angle in the lamp coordinate system; Step E3: Target angle to pixel coordinate mapping: The Micro-LED beam control module calculates the target center pixel coordinates using the center of the pixel matrix as the optical axis reference. The expression is: ; ; in, The number of the center column of the matrix. The center row number of the matrix. This represents the horizontal angular resolution corresponding to a single column of the pixel matrix. This represents the vertical angular resolution corresponding to a single row of the pixel matrix. This is the rounding function. The x-coordinate of the center pixel of the target The ordinate of the center pixel of the target; Step E4: Distance-Adaptive Masking Template Generation: The Micro-LED beam control module generates a rectangular masking template M_j for each valid target, the horizontal half-width of which is... The calculation expression is: ; in, The horizontal occlusion width of the target. For angle safety margin; The template's vertical range extends upwards from the row containing the target to above the high beam cutoff line, expressed as: ; ; ; in, This indicates the lower boundary signal of the masking template. This represents the y-coordinate of the center pixel of the j-th target. This indicates the signal at the upper boundary of the masking template. This indicates the number of rows at half the vertical height of the masking template. This indicates a safety margin for vertical upward movement. This represents the vertical angular resolution corresponding to a single row of the pixel matrix; When |c-col_j|≤n_w,j and row_top≤r≤row_bottom, the single-target occlusion template M_i(r,c)=0, and the template is turned off; otherwise, the single-target occlusion template M_j(r,c)=1, and the template is kept. Here, c represents the pixel column index, col_j represents the horizontal coordinate of the center pixel of the j-th target, r represents the pixel row index, and n_w,j represents the horizontal half-width of the occlusion template of the j-th target. Step E5: Multi-target template synthesis: The Micro-LED beam control module performs a logical AND operation on a pixel-by-pixel basis on the masking templates of all valid targets to synthesize the total masking template. M(r,c)=min(M_1(1,c),M_2(r,c),...,M_n(r,c)); Where M(r,c) represents the total occlusion template, M_n(r,c) is the occlusion template of the nth valid target, and n is the total number of valid targets; Step E6: Temporal anti-flicker processing: Asymmetric timing is used for occlusion establishment and de-occlusion. Establishment: When the frame takes effect -- the target appears, immediately output the total occlusion template M(r,c)=0 within the same refresh cycle; Unmasking: After the target disappears, the occlusion template remains for 300ms, and then gradually fades linearly within the unmasking time threshold T_fade=200ms. M_out(r,c,k)=min(M(r,c,k),M_hold(r,c,k-1)); where M_hold(r,c,k-1) represents maintaining the occlusion state of the previous frame within the holding window, M(r,c,k) is the occlusion state of the current frame, and M_out(r,c,k) is the total occlusion mask output of the current frame; Step E7: Softening the edges of the occluded area: using a Gaussian convolution kernel Output of total masking template The edges are softened to obtain the softened total masking template output. The expression is: ; in, for Gaussian convolution kernel of pixels, This is a convolution operation; Step E8: Pixel Luminance Combination: Final Pixel Luminance Matrix The calculation expression is: ; Where L_base is the mode reference luminance matrix, M_soft is the temperature derating factor, and M_soft is the anti-glare shielding factor, where M_soft∈(0,1); This represents the overall pixel translation amount; Step E9: Driving Parameter Mapping and Distribution: Mapping the final pixel brightness matrix Converted into drive circuit control quantities, and pixel-level drive control is adopted; Step E10: Oncoming / co-directional target discrimination: when the target distance change rate dd_j / dt is less than the oncoming vehicle change rate threshold, the target is determined as an oncoming vehicle; when the target distance change rate dd_j / dt falls within the co-directional preceding vehicle change rate range and the target is located ahead in the host lane, the target is determined as a co-directional preceding vehicle.

7. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 1, characterized in that: In the step 4d, the safety management module is configured to continuously monitor the system operating state, perform temperature derating, power budget allocation and hierarchical failure degradation control, as follows specifically: (1) Channel-level monitoring and protection: Overcurrent / short-circuit protection: when the current I_ch of any channel is greater than 1.5 times the protection circuit threshold I_rated and the duration reaches the set protection current time threshold, the channel is immediately turned off, and a channel fault flag is set; when faults occur in 3 or more consecutive channels, the system enters L5-level degradation; Open circuit / pixel failure detection: by using current sampling during row gating, open circuits are detected column by column, columns with channel current lower than the threshold are marked as open-circuit pixels, and the open-circuit pixels are replaced by interpolation of adjacent columns in the brightness matrix to avoid dark spots; (2) Temperature derating and power budget allocation: Temperature derating: The global luminance reduction factor is calculated based on the lamp body temperature T_sen. The expression is: When the lamp body temperature Derating temperature threshold hour, ; otherwise, ; That is, when the temperature exceeds 80°C, the global brightness decreases by 2% for every 1°C increase, down to a minimum of 50%; global brightness reduction factor. The input Micro-LED beam control module participates in brightness synthesis; when the temperature exceeds 105°C, the cooling fan starts at full speed and is further reduced in derating. Power budget allocation: Real-time estimation of the system's total power demand. The expression is: ; in, For AFS motor power, For high-side load power, Channel voltage, For channel current, This represents the PWM duty cycle of the LED channel. When total power demand Set total power threshold When needed, reduce adjustable loads from low to high according to functional priority; The priority order is: decoration / welcome light type < projection display < ADB far-field pixel < curve fill light pixel < low beam basic pixel; Power attenuation coefficient The calculation expression is: =1-( ) / ; in, Adjustable power; Degradation state machine transition conditions: the degradation level only rises and does not jump for recovery -- when the fault severity is higher than the current level, the level is immediately raised to the corresponding level; during recovery, the level is only reduced by one stage each time, and the reflow condition is that the triggering fault disappears and the duration reaches the recovery time threshold T_recover.

8. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 7, characterized in that: The safety management module performs hierarchical failure degradation control, which specifically includes the following degradation control schemes: Level L0: Normal; Trigger condition: No fault; Degradation control scheme: All functions of the system are executed normally; Level L2: Camera degradation; Trigger condition: The camera target list times out, or the camera validity flag Valid=0; Degradation control scheme: Exit the ADB mode M4, switch to low beam M1, and maintain AFS curve lighting and slope compensation; Level L3: Mechanical execution degradation; Trigger condition: Motor stall is determined, or position feedback fails; Degradation control scheme: Disconnect the motor drive output to make the lamp body return to the center via the reset spring, limit the pixel offset, and the Micro-LED pixel translation undertakes the curve beam deflection; Level L4: Communication degradation; Trigger condition: The environmental bus CAN error count is greater than the set error count threshold Bus Off, or the task bus fails; Degradation control scheme: Maintain the last valid mode for 3s and then switch to the default low beam, and the local temperature protection and low beam constant current drive continue to work; Level L5: Critical fault; Trigger condition: Overheating, i.e., lamp body temperature. >115°C, or a constant current fault or power supply abnormality in 3 or more channels; Degradation control scheme: only retain the constant current channel corresponding to the basic low beam, reduce the brightness to 50%, report the diagnostic fault code (DTC) and request the whole vehicle to switch to emergency lighting.

9. The method for coordinated control of a Micro-led digital headlight and an AFS actuator according to claim 1, characterized in that: The execution module includes an LED driving circuit, a high-side driving circuit, and a motor driving circuit. The MCU main controller drives the MicroLED combination via the LED driving circuit, drives the high-side load via the high-side driving circuit, and drives the AFS execution motor via the motor driving circuit.

Citation Information

Patent Citations

  • High beam adjusting method and system based on steering wheel angle control

    CN118514600A