Steering control method and system for automobile lighting system

Abstract

The invention discloses an automobile lighting system steering control method and system, and relates to the technical field of automobile lighting control, and the method comprises the steps: carrying out the feature matching of a real-time track of a target at the rear side of a vehicle according to a historical track mode in a target track library, so as to determine a constraint region of the real-time track; determining an interaction region between the expected steering track of the vehicle and the real-time track based on a spatial position relationship between the constraint region of the real-time track and the expected steering track of the vehicle, and identifying a typical motion mode of the real-time track according to the interaction region; determining a risk action area of the typical motion mode based on a spatial position relationship between the typical motion mode and the expected steering track of the vehicle; according to the motion trend characteristics of the real-time track in the risk action area, a steering lamp warning mode corresponding to the risk action area is triggered; accuracy and timeliness of active intelligent control of the steering lamp are achieved, and safety of the vehicle in the steering process is effectively improved.

Description

Technical Field

[0001] This invention relates to the field of automotive lighting control technology, specifically to a steering control method and system for an automotive lighting system. Background Technology

[0002] Automotive lighting control systems are crucial for enhancing vehicle safety, especially in complex urban traffic environments. Appropriate lighting signals can effectively alert surrounding vehicles and pedestrians, preventing accidents. However, current mainstream automotive turn signal control methods generally rely on manual driver triggering or simple turn signal linkage mechanisms. These methods cannot accurately adapt to the diverse and unpredictable trajectories of vehicles moving to the sides and rear in real-world traffic scenarios. This results in inaccurate timing and coverage of lighting alerts, failing to meet the precision and proactive demands of the era of intelligent driving.

[0003] While existing automotive lighting control technologies have gradually achieved automation, they typically respond only to the vehicle's own steering state or basic sensor signals, lacking in-depth analysis of the target vehicle's trajectory, especially the effective perception and accurate identification of the spatial interaction area between the vehicle's steering trajectory and the trajectory of targets to the side and rear. This deficiency makes turn signal control inflexible and lacking in initiative, failing to deliver effective warning information to other road users in a timely and accurate manner in complex traffic environments. Summary of the Invention

[0004] The purpose of this invention is to provide a steering control method and system for automotive lighting systems to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, the present invention provides a steering control method for an automotive lighting system, characterized in that it includes: The vehicle's current steering intention is used to retrieve a pre-built target trajectory library, and the real-time trajectory of the target behind and to the side of the vehicle is matched with features based on the historical trajectory patterns in the target trajectory library to determine the constraint area of ​​the real-time trajectory. Based on the spatial relationship between the constraint area of ​​the real-time trajectory and the vehicle's expected steering trajectory, the interaction area between the vehicle's expected steering trajectory and the real-time trajectory is determined, and the typical motion pattern of the real-time trajectory is identified according to the motion trend characteristics of the real-time trajectory within the interaction area. Based on the spatial relationship between typical motion patterns and the vehicle's expected steering trajectory, the risk area of ​​typical motion patterns is determined. Based on the movement trend characteristics of the real-time trajectory within the risk area, the corresponding turn signal warning mode is triggered.

[0006] Secondly, the present invention provides a steering control system for an automotive lighting system, implemented based on the method described above, comprising: The feature matching module is used to retrieve a pre-built target trajectory library based on the vehicle's current steering intention, and to perform feature matching on the real-time trajectory of the target behind and to the side of the vehicle based on the historical trajectory patterns in the target trajectory library, so as to determine the constraint area of ​​the real-time trajectory. The pattern recognition module is used to determine the interaction area between the vehicle's expected turning trajectory and the real-time trajectory based on the spatial relationship between the constraint area of ​​the real-time trajectory and the vehicle's expected turning trajectory, and to identify the typical motion pattern of the real-time trajectory based on the motion trend characteristics of the real-time trajectory within the interaction area. The position analysis module is used to determine the risk area of ​​the typical motion pattern based on the spatial positional relationship between the typical motion pattern and the vehicle's expected steering trajectory. The steering control module is used to trigger the turn signal warning mode corresponding to the risk area based on the real-time trajectory movement trend characteristics within the risk area.

[0007] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention introduces a trajectory feature matching mechanism based on a target trajectory library. By fully utilizing the spatial correlation between historical trajectory patterns and real-time target trajectories during vehicle turn signal control, the triggering of turn signals no longer depends solely on the vehicle's own turning state. This allows for accurate characterization of the potential movement range of targets to the side and rear at the initial stage of vehicle turning, thereby improving the matching degree between light prompts and real traffic conditions.

[0008] This invention performs spatial interaction analysis between the real-time trajectory constraint area and the vehicle's expected turning trajectory to further identify specific areas where trajectory interaction may occur. Based on this, it introduces a typical motion pattern recognition process, enabling the lighting control logic to differentiate for different trajectory evolution characteristics and enhance the targeting of turn signals to key interaction positions in complex traffic scenarios.

[0009] This invention constructs a risk action zone within the trajectory interaction area and drives the selection of turn signal warning modes based on the movement trend changes of the target trajectory. This enables the turn signal control to dynamically respond to changes in the target trajectory state, thereby improving the turn signal's proactive warning capability during vehicle turning and the overall intelligence level of lighting control. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0011] Figure 1 This is a flowchart of a steering control method for an automotive lighting system according to the present invention; Figure 2 This is a framework diagram of a steering control system for an automotive lighting system according to the present invention. Detailed Implementation

[0012] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided to make the description of this application more complete and comprehensive, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The drawings are merely illustrative illustrations of this application and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0013] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more exemplary embodiments. Numerous specific details are provided in the following description to give a full understanding of the exemplary embodiments disclosed in this application. However, those skilled in the art will recognize that the technical solutions disclosed in this application can be practiced with one or more specific details omitted, or other methods, components, steps, etc., can be employed. In other instances, well-known structures, methods, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the disclosure of this application.

[0014] Example 1 like Figure 1 As shown, this embodiment discloses a steering control method for an automotive lighting system, including: S101: Retrieve a pre-built target trajectory library based on the vehicle's current steering intention, and perform feature matching on the real-time trajectory of the target to the side and rear of the vehicle according to the historical trajectory patterns in the target trajectory library, so as to determine the constraint area of ​​the real-time trajectory. Specifically, the constraint region for determining the real-time trajectory includes: Construct a trajectory topology association structure based on the spatial motion association relationships of trajectory nodes in the historical trajectory patterns in the target trajectory database; Specifically, the "current steering intention of the vehicle" in this step refers to the expected steering trajectory information corresponding to the current driving state of the vehicle. It can be obtained based on the path information planned by the vehicle navigation system or the steering intention identified by the driver through changes in steering wheel angle or the action of the turn signal switch.

[0015] It should be noted that the "target trajectory database" is pre-built by collecting and analyzing historical vehicle movement trajectories in specific road areas over a long period of time, forming a database containing a large number of historical trajectory patterns. Each historical trajectory pattern in the target trajectory database consists of a series of trajectory nodes. Each trajectory node specifically contains two-dimensional spatial coordinate information (such as coordinate position in the road plane coordinate system) and a corresponding timestamp. There are clear spatial motion relationships between trajectory nodes, that is, the order of spatial position and movement direction of the trajectory nodes.

[0016] Specifically, the trajectory topology association structure is a data structure that structurally expresses the spatial topological relationships of historical trajectory patterns. Preferably, this embodiment uses a graph structure to represent the topology association structure, where each trajectory node is a vertex of the graph, the motion relationships between nodes are represented by edge connections, and the direction of the edges represents the temporal relationship of vehicle motion.

[0017] In one specific embodiment, the trajectory topological association structure can be represented as a directed graph. ,in For a set of trajectory nodes, each node Corresponding spatial location in the historical trajectory pattern; This represents the motion relationship between trajectory nodes. Any two nodes... There are directed edges between them. This indicates that the vehicle may be generated by a node. Move to node .

[0018] Preferably, when constructing the topological association structure, the historical trajectory patterns are preprocessed, specifically including trajectory node coordinate normalization and node density clustering analysis, in order to optimize the expression efficiency of the trajectory topological association structure.

[0019] Identify key trajectory nodes in the real-time trajectory based on the spatial motion change characteristics of the trajectory nodes. Specifically, real-time trajectory refers to a series of continuous motion trajectory nodes generated by a target vehicle to the side and rear of the vehicle at the current moment. These nodes are also composed of two-dimensional spatial coordinates and corresponding timestamps.

[0020] In the specific implementation process, key trajectory nodes are those trajectory nodes in the real-time trajectory that show significant changes in spatial motion characteristics and can reflect the vehicle's motion trend. The identification of these nodes is directly related to the accuracy of trajectory matching.

[0021] The key trajectory nodes for identifying real-time trajectories include: Based on the spatial motion direction change characteristics of the trajectory nodes in the real-time trajectory, the spatial motion turning regions of the trajectory nodes are extracted. Specifically, the spatial motion turning point region is the spatial location region where the motion direction of the real-time trajectory node undergoes a significant turning point or change.

[0022] In one specific embodiment, for a real-time trajectory node sequence For three consecutive nodes By calculating intermediate nodes Angle of change of direction of movement Perform identification: ;in, Indicates the degree of change in the direction of node movement, when When the angle of inflection is greater than or equal to the set directional turning angle threshold (e.g., 30°), the corresponding node will be... Key nodes marked as spatial motion transition regions.

[0023] Based on the local spatial similarity between the spatial motion transition region and the spatial motion change characteristics of trajectory nodes in the historical trajectory pattern, key trajectory nodes of the real-time trajectory are identified. It should be noted that historical trajectory patterns also contain spatial motion change nodes that are highly correlated with steering behavior. These nodes reflect the common motion turning points that occur in a large number of vehicles under the same steering scenario. By performing local spatial similarity analysis on the spatial motion turning areas in the real-time trajectory and the corresponding areas in the historical trajectory pattern, it is possible to determine whether the real-time trajectory conforms to the typical evolutionary characteristics of a certain historical trajectory pattern, thereby identifying representative key trajectory nodes.

[0024] Specifically, in one specific embodiment, the local spatial similarity relationship is determined as follows: First, extract the set of trajectory nodes corresponding to the marked spatial motion transition regions from the real-time trajectory, denoted as set. ; Simultaneously, a historical trajectory pattern corresponding to the current turning intention is selected from the target trajectory library, and the set of spatial motion change nodes in the historical trajectory pattern is extracted, denoted as set. .

[0025] Then, for the set With sets Similarity calculations are performed on the local spatial distribution relationships between them.

[0026] Preferably, a dynamic time warping method is used to align and match the spatial position sequences of the two sets of nodes to measure the consistency between the real-time trajectory and the historical trajectory pattern in terms of spatial motion change location. This similarity calculation is based on the cumulative value of spatial distance between nodes. When the cumulative distance after matching meets the preset similarity threshold condition, it is considered that the real-time trajectory and the historical trajectory pattern have consistent spatial motion change characteristics in that area.

[0027] In some specific embodiments, when a spatial motion turning point in the real-time trajectory matches the corresponding node in the historical trajectory pattern to a high degree in terms of spatial location and direction of change, the real-time trajectory node is identified as a key trajectory node for subsequent trajectory topology matching analysis.

[0028] The constraint region of the real-time trajectory is determined based on the spatial topological matching relationship between key trajectory nodes and the trajectory topological association structure. Specifically, the spatial topology matching relationship is used to describe the corresponding positional relationship of key nodes of the real-time trajectory in the topology of the historical trajectory pattern. Through this matching relationship, the possible evolution path range of the real-time trajectory in the historical trajectory pattern can be determined, thereby limiting the reasonable movement boundary of the real-time trajectory in space, i.e., the constraint region.

[0029] In one specific embodiment, spatial topology matching is implemented as follows: First, the key trajectory nodes identified in the real-time trajectory are connected in chronological order to form a local topology of the real-time trajectory, denoted as the topology. ,in, For the set of key trajectory nodes, This represents the temporal connection relationships between nodes.

[0030] Meanwhile, the trajectory topology association structure corresponding to the historical trajectory pattern in the target trajectory library is denoted as... ,in, A set of historical trajectory nodes. This refers to the motion relationships between historical trajectory nodes.

[0031] Then, regarding the topology and The connection relationships between them are matched and analyzed to calculate the matching ratio between the connection relationships of key nodes in the real-time trajectory and the topological connection relationships of the historical trajectory. This matching ratio reflects the degree of consistency between the spatial evolution sequence of the real-time trajectory and the historical trajectory pattern.

[0032] In some specific embodiments, when the above-mentioned topology matching ratio meets the preset threshold condition, it is considered that the real-time trajectory is likely to evolve along the spatial path corresponding to the historical trajectory pattern, and the spatial area covered by the historical trajectory pattern is determined as the constraint area of ​​the real-time trajectory.

[0033] S102: Based on the spatial relationship between the constraint area of ​​the real-time trajectory and the vehicle's expected steering trajectory, determine the interaction area between the vehicle's expected steering trajectory and the real-time trajectory, and identify the typical motion pattern of the real-time trajectory according to the motion trend characteristics of the real-time trajectory within the interaction area. Specifically, determining the interaction area between the vehicle's expected steering trajectory and its real-time trajectory includes: Based on the spatial overlap characteristics between the constraint area of ​​the real-time trajectory and the expected turning trajectory of the vehicle, candidate regions for trajectory interaction are determined. Specifically, in this step, the vehicle's expected steering trajectory refers to the spatial motion path planned and generated by the vehicle based on its current steering intention, used to determine the spatial area the vehicle will traverse in the future. The vehicle's expected steering trajectory is provided by the navigation system or the vehicle motion planning module, and is specifically represented as a sequence of trajectory points consisting of a series of continuous two-dimensional spatial position coordinates.

[0034] The constraint region of the real-time trajectory is the spatial range in which the target vehicle on the side and rear can move, as determined by the analysis of S101. The constraint region is represented by a polygonal region composed of spatial coordinates.

[0035] Specifically, the candidate regions for trajectory interaction are determined as follows: First, the real-time trajectory constraint area and the vehicle's expected turning trajectory are overlaid and analyzed in the same two-dimensional spatial coordinate system. Then, spatial geometric operations (such as spatial intersection algorithms) are used to calculate the overlapping area between the two. When an overlapping area exists, it is defined as a trajectory interaction candidate area, that is, the area where the real-time trajectory and the vehicle's expected turning trajectory may interact spatially.

[0036] In one specific embodiment, the spatial intersection algorithm can be implemented based on the classic polygon intersection calculation method: For example, the real-time trajectory constraint region can be represented as a polygonal region. The vehicle's expected steering trajectory is represented as a polygonal path region. Then the candidate regions for trajectory interaction between the two. It can be represented as: The symbol “∩” represents the intersection operation of the two regions. This calculation is used to obtain the set of boundary points of the trajectory interaction candidate region, thereby defining the specific spatial range of the trajectory interaction candidate region.

[0037] Based on the local spatial motion change characteristics of the real-time trajectory within the trajectory interaction candidate area, the trajectory spatial coupling region is determined; Specifically, the trajectory spatial coupling region is the area in the trajectory interaction candidate region where the local motion trend of the real-time trajectory target vehicle and the local motion trend of the vehicle's expected turning trajectory are highly similar. This region can more accurately characterize the interaction risk of the two trajectories.

[0038] The determination of the trajectory spatial coupling region includes: Identify the regions of motion trend change of trajectory nodes based on the spatial motion direction change characteristics of real-time trajectory nodes within the trajectory interaction candidate region. Specifically, the trajectory node motion trend change area represents the area where the motion trend of the real-time trajectory target vehicle may change.

[0039] In one specific embodiment, the method for identifying regions of change in motion trends includes: First, extract the set of real-time trajectory nodes within the trajectory interaction candidate area. , where each node Having the direction and angle of motion ; Then, analyze the changes in the motion direction of consecutive nodes and calculate the difference in motion direction between adjacent nodes. When the difference in the direction of motion When a preset threshold condition is met (e.g., greater than or equal to 25°), the corresponding node position is identified as a key node in the area of ​​change in motion trend, thereby obtaining the location area where the real-time trajectory target vehicle may have a change in motion trend.

[0040] Based on the local spatial similarity between the region of motion trend change and the local spatial motion trend of the vehicle's expected steering trajectory, the trajectory spatial coupling region is determined. Specifically, the determination of the trajectory spatial coupling region aims to identify the location region within the trajectory interaction candidate region where there is a consistency in spatial motion trend between the real-time trajectory target vehicle and the expected turning trajectory of the vehicle itself, that is, the local area where both vehicles may drive into together, so as to better clarify the spatial distribution of trajectory interaction risk.

[0041] In one specific embodiment, the process of determining the trajectory spatial coupling region specifically includes: First, extract the set of nodes corresponding to the regions of change in the real-time trajectory movement trend, denoted as set. Simultaneously, extract the set of trajectory nodes corresponding to the candidate regions of trajectory interaction in the vehicle's expected steering trajectory, denoted as set. ; Secondly, calculate the similarity of local spatial motion trends between the real-time trajectory node set U and the vehicle's expected turning trajectory node set V; Specifically, the calculation method for the similarity of local spatial motion trends is as follows: The local motion trends of real-time trajectory nodes (such as the motion direction and velocity vector between nodes) are compared point by point with the corresponding local motion trends of the vehicle's expected turning trajectory nodes. Quantitative analysis is performed using vector angle or cosine similarity calculation methods. For example, nodes in a real-time trajectory and the expected steering trajectory node of the vehicle The motion trend vectors are respectively denoted as and The cosine similarity of their motion trends is calculated as follows: When the similarity value When a preset threshold condition is met (e.g., greater than or equal to 0.9), the node is judged. and It exhibits high consistency in local motion trends, and this location region is included in the trajectory space coupling region; Finally, spatial clustering analysis is performed on all node positions that meet the conditions obtained from the above similarity analysis. Preferably, a density clustering algorithm (such as the DBSCAN algorithm) is used to determine the spatial range of the trajectory spatial coupling region, forming the specific spatial boundary of the trajectory spatial coupling region.

[0042] Based on the local spatial location matching characteristics between the trajectory spatial coupling region and the vehicle's expected steering trajectory, the interaction region between the vehicle's expected steering trajectory and the real-time trajectory is determined. Specifically, the interaction area is a more precise spatial area that further clarifies the actual spatial overlap or interaction between the trajectories of the two parties. It is a more refined spatial area obtained by further refining the trajectory spatial coupling area.

[0043] Specifically, methods for determining local spatial location matching features include: First, based on the trajectory space coupling region, a set of trajectory nodes that overlap with or are adjacent to the trajectory space coupling region are extracted from the vehicle's expected turning trajectory. Then, the nearest spatial distance between the set of nodes in the real-time trajectory spatial coupling region and the set of nodes in the vehicle's expected turning trajectory is calculated. For example, the Euclidean distance between the real-time trajectory nodes and the vehicle's expected trajectory nodes within the coupling region is calculated: ; All real-time trajectory node positions that meet the spatial distance condition (e.g., less than or equal to 1 meter) and the vehicle's expected turning trajectory node positions are spatially merged to form a trajectory interaction area; The resulting interactive area represents the actual location area where the vehicle may interact spatially with or even collide with a target vehicle to the side and rear, thus providing the spatial basis for the accurate triggering of the subsequent turn signal warning mode.

[0044] S103: Based on the spatial relationship between typical motion patterns and the vehicle's expected steering trajectory, determine the risk area of ​​typical motion patterns; Specifically, determining the risk area of ​​a typical movement pattern includes: Based on the spatial motion distribution characteristics of trajectory nodes in typical motion patterns, determine the spatially unstable regions of trajectory nodes; Specifically, typical motion patterns are classification representations of typical motion states of real-time trajectories within the interaction area, such as lane-changing patterns, acceleration patterns, or deceleration patterns. This embodiment preferably employs a machine learning classification algorithm to classify the motion trend characteristics of real-time trajectories within the interaction area, thereby identifying typical motion patterns. In one specific embodiment, a typical motion pattern classification method is as follows: First, extract multiple motion feature parameters of the real-time trajectory within the interactive area, such as the rate of change of velocity of the trajectory node, the rate of change of motion direction of the trajectory node, and the change of lateral displacement of the trajectory node. Then, these feature parameters are used as inputs, and a pre-trained classification model (such as a support vector machine model or a random forest model) is used to classify motion patterns to obtain the corresponding typical motion patterns.

[0045] After identifying typical motion patterns, further analysis is conducted on the spatial motion distribution characteristics of trajectory nodes within these patterns to determine the spatially unstable regions of the trajectory nodes.

[0046] Specifically, the spatially unstable region is the region in which trajectory nodes exhibit high uncertainty in spatial distribution in a typical motion pattern, such as a spatial region with high spatial density of trajectory nodes and significant changes in the motion trend of trajectory nodes. In some specific embodiments, the method for determining spatially unstable regions includes: First, extract the set of spatial coordinates of trajectory nodes corresponding to typical motion patterns, denoted as set. ; Then, spatial clustering analysis methods (such as DBSCAN clustering method) are used to perform density clustering on the spatial distribution of the trajectory node set W to obtain multiple local spatial clustering regions; Next, the variance or standard deviation of the motion direction of trajectory nodes within each cluster region is calculated to measure the uncertainty of the motion trend within the region. When the variance of the motion direction of a certain cluster region is greater than or equal to a set threshold (e.g., ...), the standard deviation is determined. When the variance of the clustering region is zero, the clustering region is defined as a spatially unstable region.

[0047] The initial risk area is determined based on the spatial overlap between the unstable region and the vehicle's expected steering trajectory. Specifically, the initial risk area is a preliminary spatial region that further clarifies the interaction risk between the vehicle and the trajectory of typical motion patterns, and is determined based on the spatial positional relationship between the spatially unstable region and the vehicle's expected turning trajectory.

[0048] The determination of the initial risk area includes: The high-density distribution area of ​​trajectory nodes is determined based on the spatial dense distribution characteristics of trajectory nodes in the spatially unstable region. Specifically, a high-density distribution area of ​​trajectory nodes indicates an area with a high spatial distribution density of trajectory nodes within a spatially unstable region. Trajectory nodes in this area have a higher risk of changing motion trends. Preferably, in one specific embodiment, a kernel density estimation method (e.g., two-dimensional Gaussian kernel density estimation) is used to calculate the spatial density distribution function of the trajectory node set. The spatial density function is calculated as follows: ;in, For Gaussian kernel function, Here, represents the spatial coordinates of the trajectory nodes, and h is the bandwidth parameter. Regions with higher density are determined based on the density function value, and regions whose density values ​​meet a preset threshold condition are defined as high-density distribution regions.

[0049] The initial risk area is determined based on the overlap between the high-density distribution area and the local spatial location of the vehicle's expected turning trajectory. Specifically, in this embodiment, the high-density distribution area and the vehicle's expected turning trajectory are superimposed in the same spatial coordinate system, and the spatial geometric intersection algorithm is used to analyze the spatial positional overlap relationship between the two.

[0050] In one specific embodiment, the initial risk impact area is determined as follows: First, the high-density distribution area of ​​trajectory nodes is represented as region set M, and the spatial coverage area of ​​the vehicle's expected turning trajectory is represented as region set N; Then, a spatial intersection algorithm (such as a polygon intersection algorithm) is used to calculate the region set M and the region set N to obtain the set of overlapping regions. The specific intersection algorithm is expressed as follows: ;in, The initial risk area is the area where the high-density distribution area overlaps with the vehicle's expected turning trajectory.

[0051] Using the methods described above, this step defines the initial risk area, which can be used to further determine the spatial impact range that typical motion patterns may have on vehicle steering behavior.

[0052] Based on the local spatial motion change trend of the real-time trajectory within the initial risk area, the risk area of ​​typical motion patterns is determined; Specifically, the risk impact area is the final determined area in a typical motion pattern that has a real potential impact on the vehicle's expected steering behavior. It is the result of further detailed analysis of the initial risk impact area.

[0053] In some specific embodiments, the risk impact area is determined in the following ways: First, multiple consecutive trajectory nodes of the real-time trajectory within the initial risk area are extracted to form a node sequence set. Each node in the node sequence has position coordinates and motion vector information; Then, analyze the local motion trend changes of each node in the node sequence set P, such as calculating the trend of velocity or direction changes between nodes. Preferably, the trend analysis is performed using motion characteristics such as velocity change rate or acceleration change. For example, taking velocity change rate analysis as an example, nodes The formula for calculating the rate of change of velocity is: ;in, and The velocity values ​​are for two consecutive nodes. and The timestamp corresponding to the node; Next, based on the rate of change of velocity In addition to the trend of motion direction, the stability and risk tendency of the trajectory node sequence are analyzed; Finally, the rate of change of speed or directional trend must meet a preset risk threshold (e.g., the rate of change of speed exceeds...). Or the change in direction is greater than The nodes and adjacent areas of a typical motion pattern are defined as the risk area, forming the spatial range of the final risk area.

[0054] S104: Based on the motion trend characteristics of the real-time trajectory within the risk area, trigger the turn signal warning mode corresponding to the risk area; Specifically, the turn signal warning mode that triggers the risk area includes: Based on the spatial motion trend characteristics of the real-time trajectory within the risk area, predict the spatial deviation trend area of ​​the real-time trajectory. Specifically, the spatial deviation trend area refers to the potential spatial area within the risk area where the real-time trajectory target vehicle may deviate from its current trajectory and enter the vehicle's expected turning trajectory in the future. The spatial deviation trend region of the predicted real-time trajectory includes: Based on the continuous change characteristics of the spatial motion direction of real-time trajectory nodes within the risk area, identify regions where the spatial direction of trajectory nodes changes abruptly. Specifically, the spatial orientation change region is the region within the risk area where the motion direction of real-time trajectory nodes changes rapidly within a local range; In one specific embodiment, the method for identifying regions with abrupt changes in spatial orientation is as follows: First, extract the real-time trajectory node sequence from the risk area. The position coordinates of each node are ; Then, calculate the change in the angle of motion direction between nodes of the continuous trajectory. ;in, Representing nodes respectively and nodes The direction and angle of motion. When the change in the direction and angle of motion... When a preset threshold is reached (e.g., greater than or equal to 30°), the node will be... The spatial location is defined as a region of abrupt change in spatial direction.

[0055] Based on the discontinuous distribution characteristics of spatial locations between regions of abrupt changes in spatial direction, the spatial deviation trend region of the real-time trajectory is determined; Specifically, methods for determining spatial deviations from trend areas include: First, the location distribution characteristics of multiple directional change nodes within the spatial directional change region are analyzed, and spatial clustering or density analysis methods are used to determine whether these nodes exhibit local clustering or discontinuous distribution characteristics. For example, preferably, a spatial density estimation method (such as kernel density estimation) is used to perform density analysis on nodes with abrupt changes in spatial orientation, thereby determining the density distribution region of the nodes. When there are obvious spatial discontinuities between multiple nodes with abrupt changes in spatial orientation (e.g., the distance between nodes is greater than or equal to 2 meters), they are marked as different regions with abrupt changes in spatial orientation. Then, based on the location and movement trend of the direction change area, combined with historical trajectory pattern data and the current movement trend of the target vehicle in real time, the location area where the target vehicle may deviate in the future is predicted, thereby determining the spatial deviation trend area.

[0056] Specifically, the spatial deviation trend area is the area where the real-time trajectory target vehicle is most likely to enter the expected turning trajectory range if it continues to move with its current motion trend. Prediction can be achieved using probability statistics or trajectory prediction algorithms (such as Kalman filters or long short-term memory network models). For example, when employing a trajectory prediction model (such as a Long Short-Term Memory network), the input is motion feature data of the current and historical nodes, and the output is multiple future time points (such as future...). The spatial location sequence of predicted trajectory nodes (seconds) is used to define the spatial range of the spatial deviation from the trend area.

[0057] Based on the spatial deviation characteristics between the spatial deviation trend area and the vehicle's expected turning trajectory, the turn signal warning mode corresponding to the risk area is triggered. Specifically, the triggering of the turn signal warning mode is intended to automatically adjust the warning state of the vehicle's turn signals in real time based on the specific spatial positional relationship between the spatial deviation trend area and the vehicle's expected turning trajectory, so as to remind the target vehicle on the real-time trajectory to pay attention to the vehicle's turning behavior, thereby reducing the risk of collision. In one specific embodiment, triggering the turn signal warning mode includes the following: First, the spatial deviation trend area and the vehicle's expected steering trajectory are superimposed in the same coordinate system to calculate the minimum spatial distance between the spatial deviation trend area and the vehicle's expected steering trajectory. Specifically, the minimum spatial distance is calculated as follows: Define the set of boundary nodes of the spatial deviation trend region as The set of nodes for the vehicle's expected turning trajectory is The minimum spatial distance between the two is expressed as: ; Then, based on the minimum spatial distance mentioned above The size of the value determines whether the threshold conditions for triggering the turn signal warning mode are met, for example: For example, when the minimum spatial distance is less than or equal to a set threshold (e.g., 0.5 meters), the turn signal warning mode is triggered; Next, based on the distance relationship between the spatial deviation trend area and the vehicle's expected turning trajectory, as well as the movement trend of the target vehicle in real time, a suitable turn signal warning mode is selected. In some specific implementations, multiple turn signal warning modes can be set, for example: Normal flashing mode: The turn signals flash at a standard frequency, used when the spatial deviation from the trend area is far from the vehicle's expected turning trajectory but there may be risks (e.g., the distance is greater than 0.3 meters and less than or equal to 0.5 meters). Enhanced flashing mode: The turn signals flash at a high frequency, used in situations where the spatial deviation from the trend area is very close to the vehicle's expected turning trajectory and the risk is greater (e.g., the distance is less than or equal to 0.3 meters). Finally, the selected turn signal warning mode is automatically sent to the vehicle's lighting control unit, which executes the corresponding turn signal control operation in real time to remind the target vehicle on the real-time trajectory to pay attention to the vehicle's turning behavior, effectively reducing the potential collision risk.

[0058] Example 2 like Figure 2 As shown in the example, the parts not detailed in this embodiment are as shown in Example 1. This embodiment discloses a steering control system for an automotive lighting system, including: The feature matching module 201 is used to retrieve a pre-built target trajectory library based on the vehicle's current steering intention, and to perform feature matching on the real-time trajectory of the target behind and to the side of the vehicle based on the historical trajectory patterns in the target trajectory library, so as to determine the constraint area of ​​the real-time trajectory. The pattern recognition module 202 is used to determine the interaction area between the vehicle's expected turning trajectory and the real-time trajectory based on the spatial relationship between the constraint area of ​​the real-time trajectory and the vehicle's expected turning trajectory, and to identify the typical motion pattern of the real-time trajectory based on the motion trend characteristics of the real-time trajectory within the interaction area. The position analysis module 203 is used to determine the risk area of ​​the typical motion pattern based on the spatial positional relationship between the typical motion pattern and the expected steering trajectory of the vehicle. The steering control module 204 is used to trigger the turn signal warning mode corresponding to the risk area based on the motion trend characteristics of the real-time trajectory within the risk area.

[0059] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters, weights, and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0060] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for controlling the steering of an automotive lighting system, characterized in that, include: The vehicle's current steering intention is used to retrieve a pre-built target trajectory library, and the real-time trajectory of the target behind and to the side of the vehicle is matched with features based on the historical trajectory patterns in the target trajectory library to determine the constraint area of ​​the real-time trajectory. Based on the spatial relationship between the constraint area of ​​the real-time trajectory and the vehicle's expected steering trajectory, the interaction area between the vehicle's expected steering trajectory and the real-time trajectory is determined, and the typical motion pattern of the real-time trajectory is identified according to the motion trend characteristics of the real-time trajectory within the interaction area. Based on the spatial relationship between typical motion patterns and the vehicle's expected steering trajectory, the risk area of ​​typical motion patterns is determined. Based on the movement trend characteristics of the real-time trajectory within the risk area, the corresponding turn signal warning mode is triggered.

2. The method according to claim 1, characterized in that, The constraint region for determining the real-time trajectory includes: Construct a trajectory topology association structure based on the spatial motion association relationships of trajectory nodes in the historical trajectory patterns in the target trajectory database; Identify key trajectory nodes in the real-time trajectory based on the spatial motion change characteristics of the trajectory nodes. The constraint region of the real-time trajectory is determined based on the spatial topological matching relationship between key trajectory nodes and the trajectory topological association structure.

3. The method according to claim 2, characterized in that, The key trajectory nodes for identifying real-time trajectories include: Based on the spatial motion direction change characteristics of the trajectory nodes in the real-time trajectory, the spatial motion turning regions of the trajectory nodes are extracted. Based on the local spatial similarity between the spatial motion transition region and the spatial motion change characteristics of trajectory nodes in the historical trajectory pattern, key trajectory nodes of the real-time trajectory are identified.

4. The method according to claim 3, characterized in that, The determination of the interaction area between the vehicle's expected steering trajectory and its real-time trajectory includes: Based on the spatial overlap characteristics between the constraint area of ​​the real-time trajectory and the expected turning trajectory of the vehicle, candidate regions for trajectory interaction are determined. Based on the local spatial motion change characteristics of the real-time trajectory within the trajectory interaction candidate area, the trajectory spatial coupling region is determined; Based on the local spatial location matching characteristics between the trajectory spatial coupling region and the vehicle's expected steering trajectory, the interaction region between the vehicle's expected steering trajectory and the real-time trajectory is determined.

5. The method according to claim 4, characterized in that, The determination of the trajectory space coupling region includes: Identify the regions of motion trend change of trajectory nodes based on the spatial motion direction change characteristics of real-time trajectory nodes within the trajectory interaction candidate region. Based on the local spatial similarity between the region of motion trend change and the local spatial motion trend of the vehicle's expected steering trajectory, the trajectory spatial coupling region is determined.

6. The method according to claim 5, characterized in that, The determination of the risk area of ​​typical movement patterns includes: Based on the spatial motion distribution characteristics of trajectory nodes in typical motion patterns, determine the spatially unstable regions of trajectory nodes; The initial risk area is determined based on the spatial overlap between the unstable region and the vehicle's expected steering trajectory. Based on the local spatial motion change trend of the real-time trajectory within the initial risk area, the risk area of ​​typical motion patterns is determined.

7. The method according to claim 6, characterized in that, The determination of the initial risk impact area includes: The high-density distribution area of ​​trajectory nodes is determined based on the spatial dense distribution characteristics of trajectory nodes in the spatially unstable region. The initial risk area is determined based on the overlap between the high-density distribution area and the local spatial location of the vehicle's expected turning trajectory.

8. The method according to claim 7, characterized in that, The turn signal warning modes corresponding to the triggering and risk-affected areas include: Based on the spatial motion trend characteristics of the real-time trajectory within the risk area, predict the spatial deviation trend area of ​​the real-time trajectory. Based on the spatial deviation characteristics between the spatial deviation trend area and the vehicle's expected steering trajectory, the turn signal warning mode corresponding to the risk area is triggered.

9. The method according to claim 8, characterized in that, The spatial deviation trend region of the predicted real-time trajectory includes: Based on the continuous change characteristics of the spatial motion direction of real-time trajectory nodes within the risk area, identify regions where the spatial direction of trajectory nodes changes abruptly. Based on the discontinuous distribution characteristics of spatial locations between regions of abrupt changes in spatial direction, the spatial deviation trend region of the real-time trajectory is determined.

10. A steering control system for an automotive lighting system, implemented based on the method of any one of claims 1-9, characterized in that, include: The feature matching module is used to retrieve a pre-built target trajectory library based on the vehicle's current steering intention, and to perform feature matching on the real-time trajectory of the target behind and to the side of the vehicle based on the historical trajectory patterns in the target trajectory library, so as to determine the constraint area of ​​the real-time trajectory. The pattern recognition module is used to determine the interaction area between the vehicle's expected turning trajectory and the real-time trajectory based on the spatial relationship between the constraint area of ​​the real-time trajectory and the vehicle's expected turning trajectory, and to identify the typical motion pattern of the real-time trajectory based on the motion trend characteristics of the real-time trajectory within the interaction area. The position analysis module is used to determine the risk area of ​​the typical motion pattern based on the spatial positional relationship between the typical motion pattern and the vehicle's expected steering trajectory. The steering control module is used to trigger the turn signal warning mode corresponding to the risk area based on the real-time trajectory movement trend characteristics within the risk area.

Images