Adaptive control method for projection lights based on multi-condition filtering and vehicle status linkage

Abstract

The present invention relates to the field of intelligent vehicle lighting technology, specifically to a method for adaptively controlling projection lamps that utilizes multi-condition filtering and vehicle status linkage, comprising the following steps: extracting permissions and exception identifiers, combining wheel speed and acceleration verification, reading slow flash codes to generate instructions, and monitoring the ToF change rate to determine if it is below a threshold to execute a fast flash self-matching scheme. The present invention improves recognition specificity by screening non-interactive targets based on the dynamic changes in the three-dimensional curvature of the object contour and the contact area. Dynamic anomalies are bidirectionally verified by combining the wheel speed pulse interval and the acceleration change rate to enhance the reliability of static interaction determination. The fast flash trigger logic is compared with the current lamp state code to ensure that lighting feedback is consistent with interaction intent. A freeze judgment is made based on ground distance changes, suppressing light state switching caused by perturbations and improving projection lamp stability. Overall, through state verification, feature screening, dynamic verification, and freeze management, control accuracy and interaction consistency are improved in multiple dimensions.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent vehicle lamps, and in particular to a method for adaptively controlling a projection lamp by linking multi-condition filtering with vehicle status. Background Art

[0002] The field of smart lighting technology encompasses intelligent control and management methods for vehicle lighting systems. The core of this technology lies in adaptively adjusting lighting functions through perception, decision-making, and execution mechanisms, taking into account vehicle operating conditions, environmental changes, and driver needs. Smart lighting technology encompasses systemic aspects such as light source control, light distribution adjustment, energy optimization, environmental perception, and interactive response. Overall, smart lighting technology not only involves intelligent changes in lighting effects but also integrates with other vehicle subsystems, such as automated driving assistance systems, night vision systems, and onboard sensor systems, enabling more efficient, safe, and energy-efficient lighting control and management.

[0003] The projector light adaptive control method, which integrates multi-condition filtering and vehicle status, refers to a method for dynamically controlling the operating mode of a vehicle's projector lights by setting multiple specific filtering conditions based on vehicle driving status information and external environmental conditions. This patent covers determining specific factors such as vehicle speed, driving direction, ambient brightness, weather conditions, and road type, filtering and screening based on the set condition priority, and determining adaptive adjustment strategies for the projector light pattern, brightness, and projection area through logical judgment and combination matching. This method typically achieves refined control of the projector light's operating status by collecting vehicle sensor output data in real time and combining it with a pre-defined condition library and decision-making rules.

[0004] Existing technologies lack cross-logical verification of the tailgate's physical lock status when determining the vehicle's standby state. This reliance on a single signal can lead to missed or misjudged anomalies, easily causing control logic confusion. Traditional methods for identifying interactive objects often rely on ambient brightness or single-plane image information, lacking the ability to comprehensively discern three-dimensional contours and dynamic area changes. This can lead to non-interactive objects such as fallen objects and snow being mistakenly identified as valid interactive signals, triggering system erroneous responses. In existing interactive detection, single wheel speed or acceleration signals are susceptible to interference from ambient vibrations or fluctuating road conditions, failing to effectively cancel out external noise and affecting interaction determination accuracy. Projector light triggering often relies on simple state switching, lacking a two-way comparison between the trigger condition and the current state, resulting in state switching lag or false flickering. Regarding ground change sensing, traditional methods generally fail to detect subtle ground dynamic fluctuations. This results in frequent flashes and switching of the projector light when rainy water accumulates or the ground shakes, increasing visual distraction and reducing driving safety and user experience stability. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a projection lamp adaptive control method that is linked to multiple condition filtering and vehicle status.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for adaptively controlling a projection lamp by linking multi-condition filtering with vehicle status, comprising the following steps:

[0007] S1: Call the ACC activation status signal and the tailgate physical lock status signal to detect the tailgate lock sensor output when ACC is not activated. When the tailgate is not locked and ACC is not activated, generate the headlight control authority flag;

[0008] S2: Collect the three-dimensional coordinates of the object boundary contour in the projection area, extract the curvature distribution of the contour nodes, calculate the contact area change, and generate an object feature abnormality flag when the curvature distribution deviates from the foot characteristics and the contact area difference exceeds the threshold;

[0009] S3: Based on the vehicle light control authority flag and the object feature abnormality flag, the wheel speed pulse signal and the three-axis acceleration value of the acceleration sensor are called to calculate the wheel speed pulse interval and the acceleration amplitude change rate to generate a dynamic behavior association verification flag;

[0010] S4: reading the projector light off, slow flashing, and fast flashing state codes according to the dynamic behavior association verification identifier, comparing the slow flashing codes during interaction, and generating a projector light state maintenance instruction;

[0011] S5: Based on the projection lamp state maintenance instruction, monitor the ground distance sequence output by the laser ToF, calculate the standard deviation and the mean change rate, and when the flash is triggered and the change rate is lower than the freeze judgment threshold, execute the flash action to generate a self-matching control scheme for the projection lamp.

[0012] As a further solution of the present invention, the vehicle light control authority identifier includes the ACC inactive state, the tailgate unlocked state, and the power supply circuit conflict judgment; the object feature abnormality identifier includes the curvature distribution deviation, the contact area change difference, and the foot feature template difference; the dynamic behavior association verification identifier includes the wheel speed pulse interval time, the acceleration amplitude change rate, and the static interval judgment; the projection lamp state maintenance instruction includes a fast flash action instruction, the original state maintenance instruction, and the slow flash state comparison result; the projection lamp self-matching control scheme includes the distance sequence standard deviation, the mean change rate, and the freezing judgment threshold.

[0013] As a further solution of the present invention, the specific steps of S1 are:

[0014] S101: Acquire the vehicle's ACC activation state signal and tailgate lock state signal, extract the tailgate lock output value in the ACC inactive state based on the acquisition timestamp, and generate a tailgate lock state signal acquisition value;

[0015] S102: Based on the tailgate lock state signal collected value, the ACC power supply circuit state signal is called, and the tailgate lock state and the ACC power supply circuit signal are cross-judged. The tailgate is set to unlocked and the ACC is not activated as the abnormal baseline state, and signal combinations that meet the abnormal baseline state conditions are screened to obtain a tailgate and power supply circuit state conflict amount;

[0016] S103: Based on the conflict amount between the tailgate and the power supply circuit state, a vehicle light control determination rule is called to perform an authority identification on the abnormal reference state signal combination to generate a vehicle light control authority identification.

[0017] As a further solution of the present invention, the specific steps of S2 are:

[0018] S201: Collecting the three-dimensional coordinates of the boundary contour of the object in the projection area, extracting the curvature values ​​between the contour nodes, establishing a correspondence between the node numbers and the curvature values, and generating node curvature distribution values;

[0019] S202: Based on the node curvature distribution value, extracting node contact area data in adjacent time slices, calculating the contact area change difference between the nodes, setting a reference value according to the node curvature deviation and the area change difference, performing double screening, and screening out node combinations that meet both the deviation reference and contact difference threshold conditions to obtain the contact feature deviation;

[0020] S203: Extracting object feature attribute information according to the contact feature deviation, calling object category standard features based on the deviation node range distribution pattern, and generating an object feature anomaly identifier.

[0021] As a further solution of the present invention, the formula for calculating the difference in contact area change between nodes is specifically:

[0022] ;

[0023] in, Represents the difference in contact area between nodes, Represents the measured value of the node contact area in the current time slice, represents the contact area measurement value of the nodes in adjacent time slices, A fixed constant representing the time slice interval, Represents the measured value of node curvature distribution in the current time slice, Represents the measured value of node curvature distribution in adjacent time slices.

[0024] As a further solution of the present invention, the specific steps of S3 are:

[0025] S301: extracting wheel speed pulse signals and three-axis acceleration values ​​based on the vehicle light control authority identifier and the object feature abnormality identifier to generate a pulse time interval sequence;

[0026] S302: Based on the pulse time interval sequence, extract the time difference between adjacent pulses, select data nodes in the static interval, extract the three-axis acceleration amplitudes of the corresponding nodes, calculate the acceleration amplitude change rate, and select data nodes with a change rate lower than a reference value according to a dynamic threshold to obtain the static contact change amount;

[0027] S303: Extracting node dynamic feature information according to the static contact variation, classifying and identifying nodes according to node attribute determination rules, and generating dynamic behavior association verification identification.

[0028] As a further solution of the present invention, the acceleration amplitude change rate calculation formula is specifically:

[0029] ;

[0030] in, represents the rate of change of acceleration amplitude, Represents the current node Axis acceleration value, The corresponding reference axis, represents the uniform pulse time difference between adjacent nodes, is the dynamic weight factor of acceleration and time difference, It represents the sum of the square terms of the three-axis acceleration components and the absolute value of the product of the weighted time difference.

[0031] As a further solution of the present invention, the specific steps of S4 are:

[0032] S401: Reading the projector light off, slow flashing, and fast flashing state codes according to the dynamic behavior association verification identifier to generate a projector light state code value;

[0033] S402: Based on the projector lamp state code value, for the verification mark of the interactive state, extract the current state code, compare it with the slow flash state code, set a benchmark for the comparison consistency, filter out nodes that meet the slow flash consistency and inconsistent, distinguish between generating a fast flash action instruction and maintaining the original state instruction, and obtain an action instruction determination value;

[0034] S403: extracting the instruction type according to the action instruction determination value, classifying and outputting the control instruction content, and generating a projection lamp state maintaining instruction.

[0035] As a further solution of the present invention, the specific steps of S5 are:

[0036] S501: Based on the projection lamp state maintenance instruction, monitor the ground distance sequence output by the laser ToF sensor, extract continuous sampling points, and generate a ground distance sampling sequence;

[0037] S502: Based on the ground distance sampling sequence, identify the distance standard deviation of consecutive sampling points, extract the ground distance mean of adjacent time slices, calculate the mean change rate, call the freezing determination threshold, filter out time slices with a distance change rate lower than the freezing determination threshold, establish the time slice index that meets the freezing condition, and obtain the ground freezing change amount;

[0038] S503: Based on the ground freezing change amount and in combination with the flash trigger instruction, the nodes that meet the flash trigger condition and the freezing condition are screened, the corresponding control action content is output, and the projection lamp self-matching control solution is generated.

[0039] As a further solution of the present invention, the mean change rate calculation formula is specifically:

[0040] ;

[0041] in, Representative The time-mean rate of change, Representative The mean ground distance of the time slice, Representative The mean ground distance of the time slice, Representative Standard deviation of ground distance between consecutive sampling points in a time slice, Representative Standard deviation of ground distance between consecutive sampling points in a time slice, Representative Time slice to The absolute value of the mean difference in ground distance between time slices.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are:

[0043] In this invention, non-interactive targets are screened out based on the dynamic changes of the object's contour three-dimensional curvature and contact area, thereby improving recognition specificity, combining the wheel speed pulse interval and acceleration change rate to bidirectionally verify dynamic anomalies, and enhancing the reliability of static interaction judgment. The flash trigger logic is compared with the current light state code to ensure that light feedback and interaction intention are consistent. Figure 1 It freezes judgment based on changes in ground distance, suppresses light state switching caused by perturbations, and improves the stability of projection lamps. Through state verification, feature screening, dynamic verification and freeze management, it improves control accuracy and interaction consistency in multiple dimensions. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 Schematic diagram of the steps of the present invention. DETAILED DESCRIPTION

[0046] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0047] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0048] In the embodiments of the present invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same. The terms "of," "corresponding," and "corresponding" may sometimes be used interchangeably. It should be noted that, when the distinction is not emphasized, the meanings they convey are the same.

[0049] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0050] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0051] See also Figure 1 The projector light adaptive control method based on multi-condition filtering and vehicle status linkage includes the following steps:

[0052] S1: Calls the vehicle's ACC activation status signal and the tailgate physical lock status signal, detects the tailgate lock sensor output signal when ACC is not activated, and compares the ACC power supply circuit and the tailgate lock mechanism status to see if there is a conflict. When the tailgate is unlocked and ACC is not activated, generates a headlight control permission flag;

[0053] S2: Collect the three-dimensional coordinate set of the object boundary contour within the projection area, extract the curvature distribution pattern between contour nodes, calculate the difference in contact area change in adjacent time slices, and compare the curvature distribution pattern with the standard curvature template of the foot. When the curvature distribution deviates from the foot characteristics and the difference in contact area exceeds the set difference threshold, an object feature abnormality flag is generated;

[0054] S3: Based on the vehicle light control authority flag and the object feature abnormality flag, the pulse signal output by the wheel speed sensor and the three-axis acceleration value output by the acceleration sensor are called to calculate the wheel speed pulse interval and the acceleration amplitude change rate. The wheel speed pulse interval is determined to be in the static range and the acceleration amplitude change rate is lower than the dynamic threshold, and a dynamic behavior association verification flag is generated.

[0055] S4: Read the current off, slow flash, and fast flash state codes of the projection lamp according to the dynamic behavior association verification flag. When the verification flag is interactive, compare the current state code with the slow flash state code to see if they are consistent, trigger the fast flash action command and the maintain original state command, and generate the projection lamp state maintain command;

[0056] S5: Based on the projector lamp status maintenance instruction, monitor the ground distance sequence output by the laser ToF sensor, calculate the standard deviation and mean change rate of the distance sequence, and when the flash trigger instruction takes effect and the distance change rate is lower than the freeze judgment threshold, execute the flash action and update the projector lamp status to generate a self-matching control scheme for the projector lamp.

[0057] The vehicle light control authority identification includes ACC inactive status, tailgate unlocked status, and power supply circuit conflict judgment; the object feature abnormality identification includes curvature distribution deviation, contact area change difference, and foot feature template difference; the dynamic behavior association verification identification includes wheel speed pulse interval time, acceleration amplitude change rate, and stationary interval judgment; the projection lamp state maintenance instruction includes fast flash action instruction, original state maintenance instruction, and slow flash state comparison result; the projection lamp self-matching control scheme includes distance sequence standard deviation, mean change rate, and freeze judgment threshold.

[0058] The specific steps of S1 are:

[0059] S101: Acquire the vehicle's ACC activation state signal and tailgate lock state signal, extract the tailgate lock output value in the ACC inactive state based on the acquisition timestamp, and generate a tailgate lock state signal acquisition value;

[0060] To obtain the vehicle's ACC activation status signal and tailgate lock status signal, first read the vehicle's current ACC ignition status through the CAN bus. The signal acquisition frequency of the ACC ignition status is set to 10 times per second. The tailgate lock status signal is read from the body control module, and the sampling frequency is also synchronously set to 10 times per second. 0 in the ACC signal represents inactive and 1 represents active. 0 in the tailgate signal represents unlocked and 1 represents locked. The data acquisition system records the system timestamp for each signal acquisition. The timestamp is recorded in milliseconds (ms). For example, ACC=0 and tailgate=1 are collected at 1728391000ms. At the same time, ACC=0 and tailgate=0 are collected again at 1728391010ms. All collected data are filtered one by one. The filtering logic is: determine whether the ACC signal value is 0. If so, keep the current sampling data. Otherwise, discard it. Initial screening is achieved by traversing the entire data set. After screening, data with similar timestamps are further processed. For example, if there are multiple records within one second, the one with the largest timestamp is taken as the valid record. This is because the data with a larger timestamp is closer to the actual state change. For example, if there are records with ACC=0 at both 1728391000ms and 1728391020ms, the record at 1728391020ms is prioritized. The collected tailgate lock status signals are sorted in ascending time order and unified into a new data table, including three columns: timestamp, ACC status, and tailgate lock status. Example data include (1728391829ms, ACC=0, tailgate lock=1) and (1728391859ms, ACC=0, tailgate lock=0). This completes the collection of the actual state output value set of the tailgate lock when ACC is not activated, ensuring the accuracy of subsequent data judgment.

[0061] S102: Based on the tailgate lock state signal acquisition value, the ACC power supply circuit state signal is called, and the tailgate lock state and the ACC power supply circuit signal are cross-judged. The tailgate is unlocked and the ACC is not activated as the abnormal baseline state. Signal combinations that meet the abnormal baseline state conditions are screened to obtain the tailgate and power supply circuit state conflict amount;

[0062] Based on the tailgate lock state collection value, the ACC power supply circuit state signal of the corresponding timestamp is called for each record. The power supply circuit state signal comes from the power management module. The signal means 0 for no power supply and 1 for power supply. The maximum time error tolerance is set to 50ms when calling. This tolerance value is based on the actual communication delay analysis. When the vehicle is stationary, the sampling error usually does not exceed 30ms. In order to prevent missed detection, the tolerance is expanded to 50ms to improve the integrity of abnormality detection. For example, if the tailgate lock signal sampling time is 1728391850ms, the ACC power supply circuit signal is searched in the range of 1728391800ms to 1728391900ms. If the closest ACC power supply circuit data is found, the tailgate lock is The cross judgment standard is set as follows: if the tailgate lock state is 0 and the ACC power supply circuit state is 0, it is recorded as the abnormal baseline state; otherwise, it is the normal state. The screening process is a one-by-one matching judgment. For example, if a set of data is (1728391859ms, tailgate lock = 0, power supply circuit = 0), it is recorded as the abnormal baseline state. If a set of data is (1728391865ms, tailgate lock = 1, power supply circuit = 0) or (tailgate lock = 0, power supply circuit = 1), it is a normal state. The abnormal combination records after screening form the tailgate and power supply circuit state conflict quantity. The conflict quantity data set is sorted in ascending order by timestamp, and the number of abnormal occurrences and time are recorded to facilitate subsequent statistical analysis or control processing.

[0063] S103: Based on the conflict amount between the tailgate and the power supply circuit states, calling the headlight control determination rule, performing an authority identification on the abnormal reference state signal combination, and generating a headlight control authority identification;

[0064] According to the conflict between the tailgate and the power supply circuit status, the headlight control judgment rules are called in turn to perform permission identification processing. The specific settings of the judgment rules are as follows: First, read the current headlight control permission status. The permission value is defined as 0, 1, 2, and 3. 0 means complete prohibition, 1 means partial restriction, only the clearance lights are allowed to be turned on, 2 means normal permission, and the low beam and high beam can be turned on normally, 3 means priority permission, and the warning lights can be turned on manually beyond the normal logic. When an abnormal baseline state is detected (the tailgate is not locked and the ACC is not powered), the headlight permission of the corresponding time node is directly set to 0, and the permission change log information is recorded at the same time. The permission change record items include: original permission value, adjusted permission value, and adjustment reason (abnormal state trigger). For example, if the original headlight permission value is 2, the permission value is directly modified to 0 after the abnormality is detected, and the " Timestamp 1728391859ms, permission reduced from 2 to 0 because the tailgate is not locked and ACC is not powered. In terms of threshold setting, to prevent frequent permission changes due to short-term misjudgments, the abnormality duration threshold is set to 2000ms, that is, if an abnormal state is detected for 2 consecutive seconds, the permission change is executed. This threshold is set based on the vehicle's regular switch action time. Usually, the vehicle tailgate locking action time is about 1000ms to 1500ms. Considering communication delay and detection stability, setting it to 2000ms is more reasonable. Threshold setting example: If an abnormality is detected at 1728391859ms and 1728393860ms, it is determined that the abnormality has lasted for more than 2000ms and the permission adjustment is executed. Otherwise, no adjustment is made. Finally, a permission adjustment result table is formed for subsequent system calls or abnormality statistical analysis.

[0065] The specific steps of S2 are:

[0066] S201: Collecting the three-dimensional coordinates of the boundary contour of the object in the projection area, extracting the curvature values ​​between the contour nodes, establishing a correspondence between the node numbers and the curvature values, and generating node curvature distribution values;

[0067] Collect the three-dimensional coordinates of the boundary contour of the object in the projection area, use laser radar or structured light sensor for real-time scanning, and set the scanning frequency to 20Hz, that is, update the three-dimensional point cloud data 20 times per second. Extract the in-body contour points during the collection process, and record the contour points in the form of three-dimensional coordinates (x, y, z). The number of each contour node increases from 1 in the order of collection, such as the first sampling point is numbered 1, the second sampling point is numbered 2, and so on. At the same time, calculate the curvature value between the contour nodes. The curvature value is estimated by taking the ratio of the chord length and arc length formed by three consecutive nodes. The chord length is the straight line distance directly connecting the two end points, and the arc length is the actual measured path length through the middle node. For example, if nodes 1 and 2 The coordinates of nodes 1 and 3 are (10,10,0), (12,11,0), and (14,13,0) respectively. The chord length from node 1 to 3 is √[(14-10)²+(13-10)²]=(√25)=5m. The arc length of the intermediate path is the distance from node 1 to 2 √5 plus the distance from node 2 to 3 √8, which is about 5.7m in total. The curvature value is 5 / 5.7=0.877. After calculating the curvature of all nodes, a one-to-one correspondence table between node numbers and curvature values ​​is established. For example, node 1 corresponds to a curvature value of 0.877, node 2 corresponds to a curvature value of 0.894, and node 3 corresponds to a curvature value of 0.881. A complete set of node curvature distribution values ​​is generated, which is used for subsequent contact characteristic analysis.

[0068] S202: Based on the node curvature distribution value, the contact area data of the nodes in adjacent time slices are extracted, and the contact area change difference between the nodes is calculated. The reference values ​​are set according to the node curvature deviation and the area change difference, and double screening is performed to select the node combination that meets both the deviation reference and the contact difference threshold conditions to obtain the contact feature deviation;

[0069] The calculation formula for the difference in contact area change between nodes is:

[0070] ;

[0071] in, Represents the difference in contact area between nodes, Represents the measured value of the node contact area in the current time slice, represents the contact area measurement value of the nodes in adjacent time slices, A fixed constant representing the time slice interval, Represents the measured value of node curvature distribution in the current time slice, represents the measured value of node curvature distribution in adjacent time slices;

[0072] Parameter definition and data source:

[0073] and : Node contact area measurement value, in square millimeters (mm²), acquired through real-time monitoring by an optical sensor. The experimental data range is

[0074] ;

[0075] mm², setting mm² (current time slice), mm² (adjacent time slices);

[0076] : The time interval is fixed constant, in seconds (s), and the system defaults to s;

[0077] and : Node curvature distribution measurement value, in millimeters⁻¹ (mm⁻¹), calculated by collecting surface geometry data using a 3D laser scanner. The data range is

[0078] ;

[0079] mm⁻¹, set mm⁻¹ (current time slice), mm⁻¹ (adjacent time slices);

[0080] Calculation process:

[0081] Calculate the contact area change term:

[0082] ;

[0083] Compute the geometric mean of curvature:

[0084] ;

[0085] Synthetic dynamic change factor:

[0086] ;

[0087] Take the absolute value and calculate the mean curvature:

[0088] ;

[0089] Final result:

[0090] ;

[0091] Parameter setting basis and quantitative description:

[0092] and :The contact area is measured by a high-precision optical sensor with a measurement error of ±0.5mm²;

[0093] : The time slice interval is controlled by the system synchronous clock with an error of ±0.001s;

[0094] and :The curvature is calculated by fitting the surface equation of the 3D scanning point cloud. The formula is: , calculation error ±0.02mm⁻¹;

[0095] Result interpretation:

[0096] The results show that the dynamic difference parameter of the node contact area mm²·mm⁻¹, reflecting the combined effect of contact area change and curvature change in adjacent time slices. Compared with the preset threshold (e.g. 5.0mm²·mm⁻¹), if , then the node combination is screened as a candidate for contact feature deviation and enters the subsequent double judgment process.

[0097] S203: Extracting object feature attribute information based on the contact feature deviation, calling object category standard features based on the deviation node range distribution pattern, and generating an object feature anomaly flag;

[0098] According to the deviation of the contact feature, the object feature attribute information is extracted from the overall data. First, the number of deviation nodes, distribution range and concentrated area position are counted. The distribution range is identified by the continuous segment of the node number. For example, the nodes with continuous deviation numbers 5, 6, 7, and 8 represent local concentrated distribution. If the distribution interval is less than 20 nodes, it is defined as a local feature. If it is greater than 50 nodes, it is defined as an overall feature. When statistically distributing, the average curvature change and average area change of each deviation node are calculated at the same time. For example, the average curvature change of an object deviation node is 0.08, and the average area change is 2.5 cm². Based on these statistical features, the preset object category standard feature database is called, which defines the standard deviation characteristics of different categories of objects. For example, for flexible objects such as sponges, the curvature change is usually greater than 0.07, and the area change is greater than 2.0 cm². For rigid objects such as metal blocks, the curvature change is usually less than 0.03, and the area change is less than 1.0 cm². By comparing the statistical characteristics of the nodes with the standard characteristics, a category matching judgment is performed. If the object characteristics are consistent with the standard flexible object characteristics, an abnormal identification is generated. The identification content includes the number of deviation nodes, distribution interval, average curvature change, average area change and the corresponding matching object category. For example, the identification record is: "Abnormal identification, 52 deviation nodes, concentrated in nodes 5-56, average curvature change 0.08, average area change 2.5 cm², matching object category: flexible material", and finally the generation of object feature abnormality identification is completed.

[0099] The specific steps of S3 are:

[0100] S301: extracting wheel speed pulse signals and three-axis acceleration values ​​based on the vehicle light control authority flag and the object feature abnormality flag to generate a pulse time interval sequence;

[0101] Based on the headlight control authority identification and object feature abnormality identification, the vehicle status monitoring module first reads the latest generated headlight control authority identification list and object feature abnormality identification list, and extracts the timestamp information corresponding to the identification as the reference time for data synchronization. Then, the wheel speed pulse signal is read from the wheel speed sensor module. The pulse signal is represented by a certain number of pulses generated per revolution. Usually, the wheel speed sensor of a car generates 48 pulses per revolution. At the same time, the three-axis acceleration value is read from the vehicle inertial measurement unit. The three axes record the acceleration in the x, y, and z directions respectively. The unit is m / s², and the sampling frequency is 100Hz, that is, the acceleration is collected every second. For 100 acceleration values, the wheel speed pulse signal is recorded by recording the pulse rising edge timestamp to extract the pulse time series. For example, if the timestamps of three consecutive pulses are detected as 1728391000ms, 1728391020ms, and 1728391040ms, the pulse intervals are 20ms and 20ms respectively, and a pulse time interval sequence is generated. The three-axis acceleration value and the pulse time interval sequence are bound to the same time base for recording, and the interpolation method is used to synchronize the acceleration data to the pulse interval time point, forming a pulse time interval sequence and a three-axis acceleration sequence set based on the pulse time point.

[0102] S302: Based on the pulse time interval sequence, extract the time difference between adjacent pulses, select the data nodes in the static interval, extract the three-axis acceleration amplitudes of the corresponding nodes, calculate the acceleration amplitude change rate, and select the data nodes with a change rate lower than the reference value based on the dynamic threshold to obtain the static contact change amount;

[0103] The calculation formula for the acceleration amplitude change rate is as follows:

[0104] ;

[0105] in, represents the rate of change of acceleration amplitude, Represents the current node Axis acceleration value, The corresponding reference axis, represents the uniform pulse time difference between adjacent nodes, is the dynamic weight factor of acceleration and time difference, It represents the sum of the square terms of the three-axis acceleration components and the absolute value of the weighted time difference product;

[0106] Parameter setting basis:

[0107] Three-axis acceleration value: obtained through real-time monitoring of the acceleration sensor. Taking a certain node as an example, the x-axis acceleration is measured , y-axis , z-axis , the value comes from the filtered mean value within the sensor sampling period.

[0108] Unified pulse time difference: The time difference between adjacent nodes is calculated by pulse time interval sequence. , the values ​​are recorded by a high-precision timing module.

[0109] Dynamic weight factor: Based on historical data statistics, the product of acceleration and time difference is analyzed for noise sensitivity, and the setting , its value decreases as the sensor sampling frequency increases. When the sampling frequency is 50Hz, The experience value range is .

[0110] Calculation process of the example:

[0111] Calculate the square of the three-axis acceleration:

[0112] ;

[0113] ;

[0114] .

[0115] Calculate the weighted time difference product term:

[0116] ;

[0117] ;

[0118] .

[0119] Substitution Calculate the weighted terms:

[0120] ;

[0121] ;

[0122] .

[0123] Sum the three axis components:

[0124] .

[0125] Divide by the square root :

[0126] ;

[0127] .

[0128] Significance of numerical results:

[0129] The result shows that the acceleration amplitude change rate of the current node is After comparing its value with the dynamic threshold reference value, if it is lower than the reference value, it is determined to be a candidate node for static contact change.

[0130] S303: Extracting node dynamic feature information based on static contact variation, classifying and identifying nodes according to node attribute determination rules, and generating dynamic behavior association verification identification;

[0131] Based on the static contact variation, the dynamic feature information of the nodes is extracted. First, the time, acceleration amplitude, and change rate of each node in the static contact variation set are recorded. Then, the nodes are classified and identified based on the node attribute judgment rule. The node attribute judgment rule is set as follows: if the node acceleration amplitude is between 9.7m / s² and 9.9m / s² and the change rate is less than 0.0005, it is marked as a completely static node. If the acceleration amplitude is between 9.5m / s² and 9.7m / s² or between 9.9m / s² and 10.1m / s² and the change rate is between 0.0005 and 0.001, it is marked as a micro-motion node. If the acceleration amplitude exceeds the above range or the change rate exceeds 0.001, it is not marked as static. The threshold interval is set with reference to the Earth's gravity acceleration standard 9. 8m / s², combined with the inertial sensor drift data to determine the allowable floating range of ±0.2m / s², the attributes of the nodes are judged one by one during classification. For example, if the node acceleration amplitude is 9.75m / s² and the change rate is 0.0003, it is determined to be a completely static node. If the amplitude is 9.92m / s² and the change rate is 0.0007, it is determined to be a micro-motion node. After classification is completed, a dynamic behavior association verification mark is generated according to the classification results of each group of nodes. The mark content includes the node time range, the number of classification results, and abnormal marking information. The mark example is: "Time period 1728391000ms to 1728395000ms, the number of completely static nodes is 20, the number of micro-motion nodes is 5, and there are no abnormal nodes", and finally a dynamic behavior association verification mark is formed.

[0132] The specific steps of S4 are:

[0133] S401: Read the projector light off, slow flashing, and fast flashing state codes according to the dynamic behavior association verification identifier, and generate a projector light state code value;

[0134] According to the dynamic behavior association verification identifier, first read the most recently generated dynamic behavior association verification identifier list in the vehicle control system, extract the timestamp and classification result corresponding to each verification identifier, and then read the projection lamp working status information in the vehicle light control module. The specific reading content is the projection lamp off state code, slow flashing state code and fast flashing state code. The three state codes are set with specific values, for example, the off state code is 0x00, the slow flashing state code is 0x01, and the fast flashing state code is 0x02. The state code is extracted from the state manager to record the corresponding time period, and arranged in ascending time order. At the same time, the dynamic behavior verification is associated with the vehicle light control module. The timestamp of the verification mark is set to match the time accuracy within a 50ms error range. After the time matching is completed, the state of the projection lamp at the corresponding moment is read and encoded into a projection lamp state code value in a unified format. The unified format is eight-digit hexadecimal representation. Each projection lamp corresponds to an independent code value. For example, if at the timestamp 1728391000ms, the verification mark is the interactive state and the projection lamp state is the slow flashing state, then the slow flashing state code 0x01 is extracted as the projection lamp state code value of the time node, and finally a projection lamp state code value set is generated. Each record contains a timestamp and a corresponding code value for subsequent action instruction processing.

[0135] S402: Based on the projector lamp state code value, for the verification mark of interactive state, extract the current state code, compare it with the slow flash state code, set a benchmark for comparison consistency, filter out nodes that meet the slow flash consistency and inconsistent, distinguish between generating a fast flash action instruction and maintaining the original state instruction, and obtain the action instruction judgment value;

[0136] Based on the projector lamp status code value, the records with the verification mark of interactive state are processed. First, all time nodes with the verification mark of interactive state are filtered out, the current projector lamp status code value of the corresponding time node is extracted, the slow flashing status code value is read and set to 0x01, and the current code value is compared with the slow flashing status code value to see if they are consistent. If they are consistent, it is recorded as a slow flashing consistent node. If they are inconsistent, it is recorded as a slow flashing inconsistent node. The comparison process is to compare one by one. For example, if the current node code value is 0x01, which is equal to the slow flashing status code value 0x01, it is recorded as consistent. Otherwise, it is recorded as inconsistent. Then, it is filtered according to the comparison result. The screening criteria are: the consistent node directly sets the action instruction to maintain the original state, and the inconsistent node sets the action instruction to execute the fast flash action. Fast flash The corresponding code value for the action is 0x02. The threshold setting standard is based on the actual visual response requirements. The slow flash frequency is set to 1Hz and the fast flash frequency is set to 3Hz. The frequency difference is greater than 2Hz as the judgment benchmark value. Based on this benchmark value, the node set that needs to execute the fast flash instruction is screened out. For example, if the node timestamp is 1728392000ms, the current code value is 0x00, and the slow flash code value is 0x01, which are inconsistent, a fast flash action instruction is generated. If the node timestamp is 1728392050ms, the current code value is 0x01, which is consistent with the slow flash code value, a maintain original state instruction is generated. Finally, the action instruction judgment value set is obtained. Each record in the set contains the timestamp, current state code, comparison result and generated action instruction.

[0137] S403: extracting the instruction type according to the action instruction determination value, classifying and outputting the control instruction content, and generating a projection lamp state maintenance instruction;

[0138] According to the action instruction judgment value, the action instruction type information in each record is extracted. The action instruction type is divided into two categories: flash action instruction and maintain original state instruction. First, the action instruction judgment value set is classified. The classification standard is the action instruction code value. If the action instruction code is 0x02, it is classified as a flash action instruction. If the action instruction code is the maintain original state instruction code (with 0x01 as the standard), it is classified as the maintain original state instruction. During the classification process, the action instruction type is checked one by one, and the specific control instruction content is output according to the classification result. The control instruction content format is uniformly set to instruction type + target lamp number + execution action code. For example Note: If a record has a timestamp of 1728392100ms, the target lamp number is 01, and the action instruction type is a flash action, the control instruction output is "instruction type: flash, number: 01, action: 0x02". If the action instruction type is to maintain the original state, the control instruction output is "instruction type: maintain, number: 01, action: 0x01". On this basis, a list of projector lamp state maintenance instructions is uniformly generated for all classified control instructions. Each instruction in the list corresponds to a unique timestamp and lamp number, which facilitates the subsequent execution of control logic or recording of historical action trajectories, and finally completes the generation of projector lamp state maintenance instructions.

[0139] The specific steps of S5 are:

[0140] S501: Based on the projector lamp state maintenance instruction, monitor the ground distance sequence output by the laser ToF sensor, extract continuous sampling points, and generate a ground distance sampling sequence;

[0141] Based on the projector lamp status maintenance instruction, first read the currently valid projector lamp status maintenance instruction list in the vehicle control system, extract the instruction timestamp as the ground status monitoring reference time, and then call the laser ToF sensor to collect ground distance data at the set frequency. The ToF sensor sampling frequency is set to 100Hz, that is, the ground distance value is collected 100 times per second. Each sampling records the timestamp and ground distance value. The distance unit is millimeter (mm). During the sampling process, continuous sampling point data is extracted. Continuous sampling points are defined as adjacent sampling data with a time interval of no more than 10ms. If the current sampling point is If the time stamp is 1000ms and the next point's time stamp is within 1010ms, it is considered a continuous point. Otherwise, if the time interval is greater than 10ms, it is processed in segments, and each continuous sampling data set is recorded to generate a ground distance sampling sequence. The ground distance sampling sequence is arranged in chronological order, and each element contains a timestamp and the corresponding ground distance value. For example, the sequence elements are recorded as (1000ms, 2500mm), (1010ms, 2502mm), (1020ms, 2503mm), and so on, forming a complete ground distance sampling sequence for subsequent processing.

[0142] S502: Based on the ground distance sampling sequence, identify the distance standard deviation of consecutive sampling points, extract the ground distance mean of adjacent time slices, calculate the mean change rate, call the freezing judgment threshold, filter out the time slices with the distance change rate lower than the freezing judgment threshold, establish the time slice index that meets the freezing condition, and obtain the ground freezing change amount;

[0143] The formula for calculating the mean change rate is:

[0144] ;

[0145] in, Representative The time-mean rate of change, Representative The mean ground distance of the time slice, Representative The mean ground distance of the time slice, Representative Standard deviation of the ground distance of consecutive sampling points in a time slice, Representative Standard deviation of ground distance between consecutive sampling points in a time slice, Representative Time slice to The absolute value of the mean difference in ground distance between time slices;

[0146] The formula for calculating The mean change rate characteristic value at each moment , proceed as follows:

[0147] Data collection and mean calculation:

[0148] and They are Hedi The average ground distance of the time slice. The specific value is obtained by arithmetic averaging the distances of all sampling points in the time slice.

[0149] Set the sampling data within a certain time slice as follows: , Calculated rice, rice.

[0150] Standard deviation calculation:

[0151] and Represents the standard deviation of the ground distance of consecutive sampling points. According to the calculation formula of sample standard deviation, use the above and Calculate the data.

[0152] The calculation formula is ,in is the sample size.

[0153] The calculation is Meter, calculated similarly rice.

[0154] Absolute value of the mean difference:

[0155] , according to the definition, calculate the Time slice and The absolute value of the difference between the time slice ground distance and the mean value.

[0156] Calculated rice.

[0157] Substitute the formula into the calculation:

[0158] Substitute the above calculation results into the main formula:

[0159] ;

[0160] This result shows that the The characteristic value of the mean change rate of the ground distance at the moment is 0.087, which is a quantitative assessment of the change amplitude of the ground distance in the time slice. The low value reflects that the ground state changes little in this time slice, and is then used in the subsequent freezing judgment threshold screening step to determine whether the ground is in a frozen state.

[0161] S503: Based on the ground freezing change and the flash trigger instruction, the nodes that meet the flash trigger conditions and the freezing conditions are screened, the corresponding control action content is output, and the self-matching control solution of the projection lamp is generated;

[0162] According to the ground frozen change, combined with the flash trigger instruction list generated in the previous order, the timestamp correspondence is compared one by one, and the nodes that meet both the flash trigger condition and the frozen condition are filtered. The filtering logic is that if the timestamp of a node is within the flash trigger instruction time range and the corresponding time slice index exists in the frozen change set, then the node is judged to meet the dual conditions and the filtering results are recorded. During the screening process, if the node only meets one of the conditions, it will be eliminated and ignored. After the screening is completed, the corresponding control action content is output for the nodes that meet the conditions. The action content includes the lamp number, action type and execution time. The action type is defined as flash execution or flash delay start. The delayed start setting delay value is 200ms. The delay value is determined according to the reaction time of the vehicle's ambient light changes. Usually, the human eye perceives flicker delays between 100ms and 250ms, so the middle value of 200ms is taken as the start delay. For example, if the node timestamp is 1728393100ms and the lamp number is 02, which meets the conditions, the output control action content "Number 02, flash execution, delay 200ms" is generated. Finally, the projection lamp self-matching control solution is generated according to the action content. Each record in the solution set describes the lamp number, action instruction, delay setting and execution timestamp in detail for direct call by the subsequent vehicle projection system.

[0163] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A projector light adaptive control method based on multi-condition filtering and vehicle status linkage, characterized in that: The following steps are involved: S1: Call the ACC activation status signal and the tailgate physical lock status signal to detect the tailgate lock sensor output when ACC is not activated. When the tailgate is not locked and ACC is not activated, generate the headlight control authority flag; S2: Collect the three-dimensional coordinates of the object boundary contour in the projection area, extract the curvature distribution of the contour nodes, calculate the contact area change, and generate an object feature abnormality flag when the curvature distribution deviates from the foot characteristics and the contact area difference exceeds the threshold; S3: Based on the vehicle light control authority flag and the object feature abnormality flag, the wheel speed pulse signal and the three-axis acceleration value of the acceleration sensor are called to calculate the wheel speed pulse interval and the acceleration amplitude change rate to generate a dynamic behavior association verification flag; S4: reading the projector light off, slow flashing, and fast flashing state codes according to the dynamic behavior association verification identifier, comparing the slow flashing codes during interaction, and generating a projector light state maintenance instruction; S5: Based on the projector lamp state maintenance instruction, monitor the ground distance sequence output by the laser ToF, calculate the standard deviation and the mean change rate, and when the flash is triggered and the change rate is lower than the freeze judgment threshold, execute the flash action to generate a self-matching control solution for the projector lamp; The specific steps of S1 are: S101: Acquire the vehicle's ACC activation state signal and tailgate lock state signal, extract the tailgate lock output value in the ACC inactive state based on the acquisition timestamp, and generate a tailgate lock state signal acquisition value; S102: Based on the tailgate lock state signal collected value, the ACC power supply circuit state signal is called, and the tailgate lock state and the ACC power supply circuit signal are cross-judged. The tailgate is set to unlocked and the ACC is not activated as the abnormal baseline state, and signal combinations that meet the abnormal baseline state conditions are screened to obtain a tailgate and power supply circuit state conflict amount; S103: Based on the conflict amount between the tailgate and the power supply circuit state, a vehicle light control determination rule is called to perform an authority identification on the abnormal reference state signal combination to generate a vehicle light control authority identification.

2. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1 is characterized in that: The vehicle light control authority identifier includes the ACC inactive state, the tailgate unlocked state, and the power supply circuit conflict judgment; the object feature abnormality identifier includes the curvature distribution deviation, the contact area change difference, and the foot feature template difference; the dynamic behavior association verification identifier includes the wheel speed pulse interval time, the acceleration amplitude change rate, and the static interval judgment; the projection lamp state maintenance instruction includes the fast flash action instruction, the original state maintenance instruction, and the slow flash state comparison result; the projection lamp self-matching control scheme includes the distance sequence standard deviation, the mean change rate, and the freezing judgment threshold.

3. The projector lamp adaptive control method based on multi-condition filtering and vehicle status linkage according to claim 1, characterized in that: The specific steps of S2 are: S201: Collecting the three-dimensional coordinates of the boundary contour of the object in the projection area, extracting the curvature values ​​between the contour nodes, establishing a correspondence between the node numbers and the curvature values, and generating node curvature distribution values; S202: Based on the node curvature distribution value, extracting node contact area data in adjacent time slices, calculating the contact area change difference between the nodes, setting a reference value according to the node curvature deviation and the area change difference, performing double screening, and screening out node combinations that meet both the deviation reference and contact difference threshold conditions to obtain the contact feature deviation; S203: Extracting object feature attribute information according to the contact feature deviation, calling object category standard features based on the deviation node range distribution pattern, and generating an object feature anomaly identifier.

4. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1, characterized in that: The calculation formula for the difference in contact area change between nodes is specifically: ； in, Represents the difference in contact area between nodes, Represents the measured value of the node contact area in the current time slice, represents the contact area measurement value of the nodes in adjacent time slices, A fixed constant representing the time slice interval, Represents the measured value of node curvature distribution in the current time slice, Represents the measured value of node curvature distribution in adjacent time slices.

5. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1, characterized in that: The specific steps of S3 are: S301: extracting wheel speed pulse signals and three-axis acceleration values ​​based on the vehicle light control authority identifier and the object feature abnormality identifier to generate a pulse time interval sequence; S302: Based on the pulse time interval sequence, extract the time difference between adjacent pulses, select data nodes in the static interval, extract the three-axis acceleration amplitudes of the corresponding nodes, calculate the acceleration amplitude change rate, and select data nodes with a change rate lower than a reference value according to a dynamic threshold to obtain the static contact change amount; S303: Extracting node dynamic feature information according to the static contact variation, classifying and identifying nodes according to node attribute determination rules, and generating dynamic behavior association verification identification.

6. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1, characterized in that: The acceleration amplitude change rate calculation formula is specifically: ； in, represents the rate of change of acceleration amplitude, Represents the current node Axis acceleration value, The corresponding reference axis, represents the uniform pulse time difference between adjacent nodes, is the dynamic weight factor of acceleration and time difference, It represents the sum of the square terms of the three-axis acceleration components and the absolute value of the product of the weighted time difference.

7. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1, characterized in that: The specific steps of S4 are: S401: Reading the projector light off, slow flashing, and fast flashing state codes according to the dynamic behavior association verification identifier to generate a projector light state code value; S402: Based on the projector lamp state code value, for the verification mark of the interactive state, extract the current state code, compare it with the slow flash state code, set a benchmark for the comparison consistency, filter out nodes that meet the slow flash consistency and inconsistent, distinguish between generating a fast flash action instruction and maintaining the original state instruction, and obtain an action instruction determination value; S403: extracting the instruction type according to the action instruction determination value, classifying and outputting the control instruction content, and generating a projection lamp state maintaining instruction.

8. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1 is characterized in that: The specific steps of S5 are: S501: Based on the projection lamp state maintenance instruction, monitor the ground distance sequence output by the laser ToF sensor, extract continuous sampling points, and generate a ground distance sampling sequence; S502: Based on the ground distance sampling sequence, identify the distance standard deviation of consecutive sampling points, extract the ground distance mean of adjacent time slices, calculate the mean change rate, call the freezing determination threshold, filter out time slices with a distance change rate lower than the freezing determination threshold, establish the time slice index that meets the freezing condition, and obtain the ground freezing change amount; S503: Based on the ground freezing change amount and in combination with the flash trigger instruction, the nodes that meet the flash trigger condition and the freezing condition are screened, the corresponding control action content is output, and the projection lamp self-matching control solution is generated.

9. The projector lamp adaptive control method with multi-condition filtering and vehicle status linkage according to claim 1, characterized in that: The mean change rate calculation formula is specifically: ； in, Representative The time-mean rate of change, Representative The mean ground distance of the time slice, Representative The mean ground distance of the time slice, Representative Standard deviation of ground distance between consecutive sampling points in a time slice, Representative Standard deviation of ground distance between consecutive sampling points in a time slice, Representative Time slice to The absolute value of the mean difference in ground distance between time slices.

Images