Potential traffic safety analysis method, device, electronic device and storage medium

Abstract

The present invention relates to the technical field of traffic safety analysis, and specifically discloses a potential traffic safety analysis method, device, electronic device and storage medium, wherein the method comprises the steps of: obtaining traffic signal timing information, road feature information and traffic participant trajectory data set of a target intersection; obtaining motion trajectory information, motion parameter information and size information associated with the time series of each traffic participant according to the traffic participant trajectory data set; extracting all potential conflict event information according to the motion trajectory information, motion parameter information and size information; obtaining the spatiotemporal distribution law of the traffic conflict frequency at the target intersection; the method overcomes the limitations of the traditional accident data-based analysis method, can intuitively display the distribution law of traffic conflicts in time and space, can more comprehensively consider various factors affecting intersection safety, improves the accuracy and reliability of the analysis results, and provides a quantitative basis for improving traffic safety.

Description

Technical Field

[0001] The present application relates to the technical field of traffic safety analysis, and in particular to a potential traffic safety analysis method, device, electronic device and storage medium. Background Art

[0002] With the rapid development of road traffic and the steady growth of car ownership, traffic safety issues have always attracted much attention. Especially at signal-controlled intersections, which are key nodes in urban transportation networks, traffic safety issues are particularly important.

[0003] Traditional traffic safety analysis methods mainly rely on traffic accident data, but this method has lags and limitations, making it difficult to fully reveal the potential safety risks in the traffic system, and unable to analyze the safety performance of different time periods within the intersection signal cycle and at different locations within the intersection.

[0004] There is currently no effective technical solution to the above problems. Summary of the invention

[0005] The purpose of this application is to provide a potential traffic safety analysis method, device, electronic device and storage medium, so as to perform safety analysis of intersections by analyzing the actual behavior of traffic participants, so as to analyze the safety performance of different time periods within the intersection signal cycle and different locations within the intersection.

[0006] In a first aspect, the present application provides a potential traffic safety analysis method for analyzing traffic conflict patterns, the method comprising the following steps:

[0007] S1, obtaining traffic signal timing information, road feature information and a traffic participant trajectory data set about at least one traffic signal cycle at a target intersection;

[0008] S2. Acquire time-series motion trajectory information, motion parameter information, and size information of each traffic participant according to the traffic participant trajectory data set;

[0009] S3, extracting all potential conflict event information according to the motion trajectory information, motion parameter information and size information, wherein the potential conflict event information includes: conflict position information and conflict time information;

[0010] S4. Based on the conflict position information and the conflict time information, the road feature information and the traffic signal timing information are respectively matched to obtain the spatiotemporal distribution law of the traffic conflict frequency at the target intersection.

[0011] The potential traffic safety analysis method of the present application overcomes the limitations of traditional accident data-based analysis methods and can evaluate the safety status of intersections in real time and dynamically. By analyzing the actual behavior of traffic participants rather than relying solely on accident data that has occurred, this method can identify potential safety risks earlier, thereby improving the foresight and preventive nature of safety analysis. At the same time, this method can intuitively display the distribution patterns of traffic conflicts in time and space by combining road characteristics and traffic signal timing information, and can more comprehensively consider various factors affecting intersection safety, thereby improving the accuracy and reliability of the analysis results and providing a quantitative basis for improving traffic safety.

[0012] In the potential traffic safety analysis method, the types of potential conflict event information are divided into potential lateral conflict events and potential longitudinal conflict events, and step S3 includes:

[0013] S31, obtaining all suspected lateral conflict events in each frame and predicting the corresponding conflict time difference according to the motion trajectory information, motion parameter information and size information, and comparing the conflict time difference with a first preset threshold to filter the suspected lateral conflict events to obtain the potential lateral conflict events;

[0014] S32, obtaining all suspected longitudinal conflict events in each frame and predicting the corresponding collision time information according to the motion trajectory information, motion parameter information and size information, and comparing the collision time information with a second preset threshold value to filter the suspected longitudinal conflict events to obtain the potential longitudinal conflict events.

[0015] The potential traffic safety analysis method of the present application effectively solves the problem of extracting and classifying potential conflict event information by dividing potential conflict events into two categories: horizontal and vertical, and adopting different identification and screening methods respectively.

[0016] The potential traffic safety analysis method, wherein step S31 comprises:

[0017] S311, obtaining conflict position information of all suspected lateral conflict events in each frame according to the motion trajectory information and size information;

[0018] S312, predicting the conflict time difference of the traffic participant arriving at the conflict position information according to the conflict position information and the motion trajectory information and motion parameter information of the corresponding traffic participant;

[0019] S313: Compare the conflict time difference with a first preset threshold value to filter out the suspected lateral conflict events and obtain the potential lateral conflict events.

[0020] In this example, step S311 traverses and analyzes the spatial position, motion trajectory and size characteristics of any two traffic participants in each frame, and accurately locates the conflict position where the conflict may occur, which can be a position point or an area; step S312 is based on the acquired conflict position information, combined with the motion trajectory information and motion parameter information of the corresponding traffic participants, to accurately predict the time difference for the traffic participants among the two traffic participants who are moving toward the conflict position where the conflict may occur to arrive at the conflict position, and accurately evaluates the time it takes to arrive at the conflict position by calculating the traffic participants' movement speed, acceleration and other parameters; step S313 compares the predicted conflict time difference with a preset first threshold value, so as to screen out the real potential lateral conflict events. This screening method based on the time difference threshold can effectively filter out misjudged conflict events and improve the accuracy of recognition.

[0021] The potential traffic safety analysis method, wherein step S32 comprises:

[0022] S321, acquiring all suspected vertical conflict events in each frame according to the motion trajectory information, motion parameter information and size information;

[0023] S322, predicting corresponding collision time information according to the motion parameter information and size information of the traffic participant corresponding to the suspected longitudinal collision event;

[0024] S323: Compare the collision time information with the second preset threshold value to filter the suspected longitudinal conflict event to obtain the potential longitudinal conflict event.

[0025] The potential traffic safety analysis method, wherein the motion parameter information includes speed information and acceleration information;

[0026] Step S322 includes:

[0027] S3221, calculating collision travel information when the rear traffic participant moves to overlap with the initial position of the front traffic participant according to the speed information, acceleration information and size information of the traffic participant corresponding to the suspected longitudinal collision event;

[0028] S3222. Calculate and obtain the collision time information according to the collision travel information and the speed information and acceleration information of the corresponding traffic participant.

[0029] The potential traffic safety analysis method, wherein step S32 further comprises:

[0030] S324: Calculate collision index information according to the collision time information and the motion parameter information of the corresponding traffic participant.

[0031] The potential traffic safety analysis method, wherein step S4 comprises:

[0032] S41, establishing a first distribution regularity diagram of traffic conflict frequency with respect to time of the potential conflict event information of the target intersection based on the pairing relationship between the conflict time information and the traffic signal timing information;

[0033] S42: establishing a second distribution regularity diagram of the traffic conflict frequency with respect to the position of the potential conflict event information of the target intersection based on the pairing relationship between the conflict position information and the road feature information.

[0034] In a second aspect, the present application further provides a potential traffic safety analysis device for analyzing traffic conflict rules, the device comprising:

[0035] A data acquisition module, used to acquire traffic signal timing information, road feature information, and a data set of traffic participant trajectories for at least one traffic signal cycle at a target intersection;

[0036] A feature extraction module, used for obtaining the time-series motion trajectory information, motion parameter information and size information of each traffic participant according to the traffic participant trajectory data set;

[0037] A conflict analysis module, used to extract all potential conflict event information according to the motion trajectory information, motion parameter information and size information, wherein the potential conflict event information includes: conflict position information and conflict time information;

[0038] The conflict reconstruction module is used to respectively pair the road feature information and the traffic signal timing information based on the conflict position information and the conflict time information to obtain the spatiotemporal distribution law of the traffic conflict frequency at the target intersection.

[0039] The potential traffic safety analysis device of the present application overcomes the limitations of traditional accident data analysis devices and can evaluate the safety status of intersections in real time and dynamically. By analyzing the actual behavior of traffic participants rather than relying solely on accident data that has occurred, the device can identify potential safety risks earlier, thereby improving the foresight and preventive nature of safety analysis. At the same time, the device can intuitively display the distribution patterns of traffic conflicts in time and space by combining road characteristics and traffic signal timing information, and can more comprehensively consider various factors affecting intersection safety, thereby improving the accuracy and reliability of the analysis results and providing a quantitative basis for improving traffic safety.

[0040] In a third aspect, the present application further provides an electronic device, comprising a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps in the method provided in the first aspect are executed.

[0041] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps in the method provided in the first aspect are executed.

[0042] As can be seen from the above, the present application provides a potential traffic safety analysis method, device, electronic device and storage medium, wherein the potential traffic safety analysis method overcomes the limitations of traditional accident data-based analysis methods, and can evaluate the safety status of intersections in real time and dynamically. By analyzing the actual behavior of traffic participants rather than relying solely on accident data that has occurred, this method can identify potential safety risks earlier, thereby improving the foresight and preventive nature of safety analysis. At the same time, this method can intuitively display the distribution patterns of traffic conflicts in time and space by combining road characteristics and traffic signal timing information, and can more comprehensively consider various factors affecting intersection safety, thereby improving the accuracy and reliability of the analysis results and providing a quantitative basis for improving traffic safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 A flowchart of a potential traffic safety analysis method provided in an embodiment of the present application.

[0044] Figure 2 Schematic diagram of potential horizontal conflict events.

[0045] Figure 3 Schematic diagram of potential vertical conflict events.

[0046] Figure 4 Schematic diagram of a first distribution pattern diagram in some embodiments.

[0047] Figure 5 Schematic diagram of the second distribution law diagram in some embodiments.

[0048] Figure 6 Graph showing the temporal and spatial distribution patterns in some embodiments.

[0049] Figure 7 A schematic diagram of the structure of a potential traffic safety analysis device provided in an embodiment of the present application.

[0050] Figure 8 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0051] Figure numerals: 201, data acquisition module; 202, feature extraction module; 203, conflict analysis module; 204, conflict reconstruction module; 301, processor; 302, memory; 303, communication bus. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0053] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0054] Potential traffic safety analysis mainly relies on traffic accident data for evaluation, and this method has obvious technical limitations. Specifically, accident data has a lag and cannot reflect the current operating status of the traffic system in a timely manner; at the same time, the accident data collection process lacks systematicity, making it difficult to fully reveal the potential safety risks at different time periods and locations within the intersection. As a result, the traffic safety analysis results are significantly different from the actual situation, affecting the accuracy and timeliness of traffic safety assessments.

[0055] Taking a typical signal-controlled intersection as an example, the traffic signal timing scheme will directly affect the behavior patterns of traffic participants. In a signal cycle, at different phase transition moments, traffic participants may suddenly accelerate, decelerate or turn, and these behaviors interact with the road geometry to form a complex traffic flow pattern. Traditional analysis methods based on accident data cannot capture these micro-behavioral characteristics, nor can they identify the potential conflict risks caused by these behaviors. For example, at the moment the traffic light changes, right-turning vehicles and straight-moving non-motor vehicles may intertwine, and this potential conflict pattern is difficult to reflect in accident data.

[0056] This analysis method leads to serious technical defects in traffic safety assessment. First, it is impossible to achieve real-time monitoring and early warning of traffic conflicts, which increases the possibility of traffic accidents. Second, due to the lack of accurate evaluation of the safety performance of different areas in the intersection, it is difficult to formulate targeted improvement measures. Third, there is a lack of quantitative basis for the safety evaluation of traffic signal timing schemes, which affects the effect of traffic control. These technical problems urgently need a new analysis method to solve.

[0057] This application has conducted in-depth thinking and exploration on the problems existing in potential traffic safety analysis.

[0058] First, please refer to Figure 1-Figure 6 Some embodiments of the present application provide a potential traffic safety analysis method for analyzing traffic conflict rules, the method comprising the following steps:

[0059] S1, obtaining traffic signal timing information, road feature information and a traffic participant trajectory data set about at least one traffic signal cycle at a target intersection;

[0060] S2. Acquire the time-series motion trajectory information, motion parameter information and size information of each traffic participant according to the traffic participant trajectory data set;

[0061] S3, extracting all potential conflict event information according to the motion trajectory information, motion parameter information and size information, the potential conflict event information including: conflict position information and conflict time information;

[0062] S4. Based on the conflict position information and the conflict time information, the road feature information and the traffic signal timing information are matched respectively to obtain the spatiotemporal distribution law of the traffic conflict frequency at the target intersection.

[0063] Specifically, the target intersection is a traffic intersection that requires traffic safety analysis, is equipped with traffic lights, and has bifurcated roads.

[0064] More specifically, traffic signal timing information refers to the control scheme of the traffic lights at the intersection, including the green light time, yellow light time and full red light time in each direction of the target intersection.

[0065] More specifically, road characteristic information refers to the geometric layout and traffic facility information of the intersection, including the number of lanes, lane width, road dividing lines, the setting of turn lanes, etc., which can be obtained by field measurement or geographic information system data.

[0066] More specifically, the traffic participant trajectory dataset refers to data information used to describe and record the motion trajectories of all traffic participants in an intersection within a certain period of time. It may include bounding boxes of different traffic participants at different times or frames. The motion trajectories, motion parameters and other data of the traffic participants can be analyzed by combining the bounding boxes of the traffic participants at different times / frames. Therefore, the dataset can reflect the time-series motion trajectory information and size information of all traffic participants. The traffic participant trajectory dataset can be obtained by collecting and performing data recognition using equipment such as video surveillance systems or radar detectors.

[0067] More specifically, motion trajectory information refers to the movement path of traffic participants in time and space, which can be represented by a continuous sequence of space-time coordinate points.

[0068] More specifically, motion parameter information refers to physical quantities that describe the motion state of traffic participants, such as speed and acceleration, which can be obtained by differential calculation of motion trajectory information.

[0069] More specifically, the size information refers to the space occupied by the traffic participant, such as the length, width, and height of the vehicle, which can be determined through image processing or preset traffic participant type parameters.

[0070] More specifically, traffic participants are traffic objects moving at the target intersection, such as pedestrians, cars, bicycles, tricycles, trucks, motorcycles, buses, etc.

[0071] More specifically, potential conflict event information refers to the time and location where a traffic conflict may occur, which can be obtained by analyzing the movement trajectories and parameter predictions of traffic participants. Traffic conflict refers to the traffic accident itself and phenomena that may induce traffic accidents.

[0072] More specifically, step S1 preferably involves collecting data by taking aerial photos of the intersection through a swarm of drones, with the shooting angle trying to capture the entire intersection and more than 50 meters upstream of the stop line of each entrance lane of the intersection, so as to obtain continuous video data, and obtain road feature information and a traffic participant trajectory data set for at least one traffic signal cycle based on the video data; wherein, traffic signal timing information can also be collected and acquired through a swarm of drones, or it can be collected using timing data recorded by a signal controller; the traffic participant trajectory data set for at least one traffic signal cycle is collected so that the subsequent analysis process can establish a spatiotemporal distribution law of the traffic conflict frequency within a cycle. In an embodiment of the present application, it is preferred to obtain a traffic participant trajectory data set for at least 100 traffic signal cycles to comprehensively analyze the spatiotemporal distribution law of the traffic conflict frequency; wherein, the traffic signal cycle refers to a cycle of the traffic signal timing information; it is also worth mentioning that the potential traffic safety analysis method of the embodiment of the present application can analyze the spatiotemporal distribution law of the traffic conflict frequency corresponding to different time periods by acquiring data from different time periods (such as rush hour, noon, rush hour, late night).

[0073] More specifically, based on the foregoing content, it can be known that the traffic participant trajectory dataset records the data information of the motion trajectories of all traffic participants in the intersection within a certain period of time, which can be obtained by analyzing the video data collected by the drone swarm through a computer vision recognition algorithm, such as analyzing the video data through the Yolo algorithm to obtain the traffic participants in each frame of the video and form a suitable bounding box. Step S2 analyzes the bounding boxes of these traffic participants to obtain the traffic participants' motion trajectory information, motion parameter information and size information associated with the time series; it should be noted that in order to distinguish different traffic participants, the motion trajectory information, motion parameter information and size information are paired with corresponding traffic participant ID information or number information.

[0074] More specifically, after the motion trajectory information, motion parameter information and size information are extracted in step S2, the motion state of the traffic participants after a period of time can be predicted based on the motion trajectory information and motion parameter information of the traffic participants in different frames, the type of traffic participants can be analyzed based on the size information, and the motion boundary of the traffic participants after a period of time can be predicted. Based on these predicted state analyses, potential conflict events can be analyzed; therefore, step S3 predicts the time and location of possible traffic conflicts by comparing these data information of different traffic participants, thereby identifying potential conflict events. This step takes into account the motion state and space occupancy of traffic participants, and can more accurately capture the potential safety risks at different time nodes within the traffic signal cycle.

[0075] More specifically, after obtaining all the potential conflict event information representing the potential safety risks at different time nodes within the traffic signal cycle, step S4 combines this potential conflict event information with the road feature information and the traffic signal timing information to analyze the spatiotemporal distribution of traffic conflict frequencies; this step reveals the relationship between traffic conflicts and intersection environments by considering road geometric characteristics and signal control strategies, providing a more targeted basis for improving traffic safety.

[0076] The potential traffic safety analysis method of the embodiment of the present application overcomes the limitations of the traditional accident data-based analysis method, and can evaluate the safety status of intersections in real time and dynamically. By analyzing the actual behavior of traffic participants rather than relying solely on accident data that has occurred, the method can identify potential safety risks earlier, thereby improving the foresight and preventive nature of safety analysis. At the same time, the method can intuitively display the distribution patterns of traffic conflicts in time and space by combining road characteristics and traffic signal timing information, and can more comprehensively consider various factors affecting intersection safety, thereby improving the accuracy and reliability of the analysis results and providing a quantitative basis for improving traffic safety.

[0077] In some preferred implementations, the types of potential conflict event information are divided into potential horizontal conflict events and potential vertical conflict events. Step S3 includes:

[0078] S31, obtaining all suspected lateral conflict events in each frame and predicting the corresponding conflict time difference according to the motion trajectory information, motion parameter information and size information, and comparing the conflict time difference with a first preset threshold value to filter out suspected lateral conflict events to obtain potential lateral conflict events;

[0079] S32, obtaining all suspected longitudinal conflict events in each frame and predicting the corresponding collision time information according to the motion trajectory information, motion parameter information and size information, and comparing the collision time information with the second preset threshold to filter out suspected longitudinal conflict events to obtain potential longitudinal conflict events.

[0080] Specifically, the potential traffic safety analysis method of the embodiment of the present application effectively solves the problem of extracting and classifying potential conflict event information by dividing potential conflict events into two categories: horizontal and vertical, and adopting different identification and screening methods respectively.

[0081] More specifically, suspected lateral conflict events and suspected longitudinal conflict events are events that are preliminarily analyzed to be likely to result in traffic conflicts in a lateral or longitudinal relationship, respectively.

[0082] More specifically, the potential traffic safety analysis method of the embodiment of the present application uses motion trajectory information, motion parameter information and size information as basic data, and realizes the screening of suspected conflict events by predicting the conflict time difference or collision time information and comparing it with the preset threshold. It can not only comprehensively capture potential conflict events, but also improve the accuracy of recognition through threshold screening, thereby providing reliable data support for subsequent traffic safety analysis.

[0083] More specifically, in the embodiments of the present application, the classification and extraction process of potential conflict event information can be implemented in a variety of ways. A lateral conflict event represents a lateral collision conflict event of a traffic participant, and its future position can be predicted based on the movement trajectory of the traffic participant, and a possible intersection can be calculated. The conflict time difference can be obtained by calculating the time difference for the traffic participant to reach the intersection from the current position. The first preset threshold can be set according to the specific situation of the intersection, for example, it can be set to 1-5 seconds.

[0084] More specifically, a longitudinal collision event represents a rear-end collision collision event of a traffic participant, which can predict the collision time by calculating the rate of change of the distance between the front and rear vehicles. The collision time information can be calculated by considering the current speed, acceleration and safety distance of the vehicle. The second preset threshold can also be set according to the actual situation, and can usually be set to 1-5 seconds.

[0085] More specifically, these two types of conflict event identification methods cooperate with each other to comprehensively cover potential conflict situations that may occur at intersections. Lateral conflict events are mainly aimed at traffic participants on the intersecting paths, while longitudinal conflict events focus on traffic participants on the same path. Through this classification method, the potential traffic safety analysis method of the embodiment of the present application can more accurately identify and analyze different types of traffic conflicts.

[0086] It should be noted that both the potential horizontal conflict event and the potential vertical conflict event include conflict position information and conflict time information.

[0087] More specifically, by dividing conflict events into two categories, horizontal and vertical, the potential traffic safety analysis method of the embodiment of the present application can more accurately capture different types of traffic conflicts, making subsequent safety analysis more targeted and able to formulate more accurate preventive measures for different types of traffic conflicts. Secondly, compared with the traditional method that only relies on position information, the use of motion trajectory information, motion parameter information and size information as basic data can more comprehensively consider the dynamic characteristics of traffic participants, thereby improving the accuracy of conflict event identification. In addition, by introducing preset thresholds to screen potential conflict events, the false alarm rate is effectively reduced, which can not only capture potential dangerous situations, but also avoid generating a large amount of meaningless misjudgment of potential conflict event information due to excessive sensitivity.

[0088] More specifically, the process of obtaining suspected lateral conflict events in step S31 and suspected longitudinal conflict events in step S32 is actually to traverse and analyze whether any two traffic participants in each frame are suspected of having a conflict relationship in terms of the lateral relationship and the longitudinal relationship. However, in reality, these traffic conflicts will not occur between pedestrians. Therefore, in some preferred embodiments, step S3 also includes the following steps executed before step S31:

[0089] S30. Determine the type of traffic participants based on their size information, and shield conflict analysis between pedestrians according to their types, so that step S31 skips the process of simultaneously selecting two pedestrians to analyze whether there is a suspected lateral conflict event, and step S32 skips the process of simultaneously selecting two pedestrians to analyze whether there is a suspected longitudinal conflict event.

[0090] Specifically, this processing method can skip the process of traffic conflict analysis between pedestrians, eliminate invalid analysis processes, and effectively improve the analysis efficiency and analysis accuracy of the potential traffic safety analysis method of the embodiment of the present application.

[0091] In some other real-time methods, step S31 may also be to screen out suspected lateral conflict events between pedestrians to eliminate these invalid traffic conflicts after obtaining all suspected lateral conflict events in each frame, and step S32 may also be to screen out suspected longitudinal conflict events between pedestrians to eliminate these invalid traffic conflicts after obtaining all suspected longitudinal conflict events in each frame.

[0092] In some preferred embodiments, step S31 includes:

[0093] S311, obtaining conflict position information of all suspected lateral conflict events in each frame according to the motion trajectory information and size information;

[0094] S312, predicting the conflict time difference of the traffic participants arriving at the conflict position information according to the conflict position information and the motion trajectory information and motion parameter information of the corresponding traffic participants;

[0095] S313: Compare the conflict time difference with the first preset threshold value to filter suspected lateral conflict events and obtain potential lateral conflict events.

[0096] Specifically, step S311 traverses and analyzes the spatial positions, motion trajectories and size characteristics of any two traffic participants in each frame, and accurately locates the conflict position where the conflict may occur, which can be a position point or an area; step S312 is based on the acquired conflict position information, combined with the motion trajectory information and motion parameter information of the corresponding traffic participants, to accurately predict the time difference for the traffic participants among the two traffic participants who are moving towards the conflict position where the conflict may occur to arrive at the conflict position, and accurately evaluates the time it takes to arrive at the conflict position by calculating the traffic participants' movement speed, acceleration and other parameters; step S313 compares the predicted conflict time difference with a preset first threshold value, so as to screen out the real potential lateral conflict events. This screening method based on the time difference threshold can effectively filter out misjudged conflict events and improve the accuracy of recognition.

[0097] It should be noted that the suspected lateral conflict event extracted in step S311 also includes conflict time information, so that the potential lateral conflict event finally determined includes both conflict position information and conflict time information.

[0098] More specifically, the potential traffic safety analysis method of the embodiment of the present application adopts a post-intrusion time algorithm to analyze and implement steps S311 to S313, such as Figure 2 As shown, taking the lateral collision of vehicles as an example, step S311 can analyze the intersection of the trajectories of the two cars as traffic participants based on the motion trajectory information and size information ( Figure 2 a in the figure), taking this as the conflict position information of the suspected lateral conflict event, and defining the time of the frame as the conflict time information; wherein, the trajectory intersection point is an overlapping position point of the motion area generated by two traffic participants based on the motion trajectory information, and the process of obtaining the trajectory intersection point can be: widening the motion trajectory of one of the vehicles based on the vehicle size to obtain the motion area of ​​the vehicle, and then analyzing whether the motion area overlaps with the current position of other vehicles, and if so, taking the position of a reference point of the overlapping part as the conflict position information, and the vehicle constitutes the rear invading vehicle of the suspected lateral conflict event; then step S312 predicts the time required for the rear invading vehicle to reach the conflict position information according to the motion trajectory and motion parameter information of the rear invading vehicle as the conflict time difference, and finally step 313 compares the size relationship between the conflict time difference and the first preset threshold to determine whether the suspected lateral conflict event belongs to a potential lateral conflict event.

[0099] More specifically, in the embodiment of the present application, the first preset threshold is preferably set to 1.5 seconds, that is, a suspected lateral conflict event with a conflict time difference of less than 1.5 seconds is a potential lateral conflict event.

[0100] In some preferred embodiments, step S32 includes:

[0101] S321, acquiring all suspected vertical conflict events in each frame according to the motion trajectory information, motion parameter information and size information;

[0102] S322, predicting corresponding collision time information according to motion parameter information and size information of the traffic participant corresponding to the suspected longitudinal collision event;

[0103] S323: Compare the collision time information and the second preset threshold value to filter out suspected longitudinal conflict events and obtain potential longitudinal conflict events.

[0104] Specifically, the potential traffic safety analysis method of the embodiment of the present application can not only detect potential longitudinal conflict risks in a timely manner but also avoid misjudgment by performing multi-step screening and prediction of longitudinal conflict events. It uses multi-dimensional data such as motion trajectory, motion parameters and size information, combined with preset thresholds for judgment, making the identification of longitudinal conflict events more accurate and reliable.

[0105] More specifically, similar to step S311, step S321 can combine the motion trajectory information and size information to determine the movement area of ​​the traffic participant, and then combine the motion parameter information to predict the position of the traffic participant in the movement area in the following period of time. Therefore, step S321 can traverse and analyze the degree of overlap of the movement areas of any two traffic participants in each frame and the changing trend of the position distance in the following period of time to determine whether the two may have a longitudinal conflict (rear-end collision) and constitute a suspected longitudinal conflict event.

[0106] More specifically, in order to further simplify the analysis process, in an embodiment of the present application, steps S321-S323 are also preferably designed and implemented using post-intrusion time thinking, that is, step S321 analyzes whether the movement area of ​​the rear traffic participant overlaps with the current position of other traffic participants moving in the same direction and determines a suspected longitudinal conflict event in combination with the motion parameter information, and then step S322 predicts the corresponding collision time information based on the motion parameter information and size information of the rear traffic participant and the size information of the front traffic participant (equivalent to predicting the collision time information corresponding to the state where the rear traffic participant moves to the corresponding frame and contacts the front traffic participant).

[0107] More specifically, the implementation of step S322 can be based on a physical motion model, for example, a uniform motion or uniform acceleration motion model can be used to calculate the potential collision time in combination with the current position, speed, acceleration and size information of the traffic participant. The specific calculation method can be selected according to the actual situation, for example, the braking distance, reaction time and other factors of the traffic participant can be considered.

[0108] More specifically, step S323 obtains potential longitudinal conflict events by screening suspected longitudinal conflict events by comparing the collision time information and the second preset threshold value. This threshold-based screening method can effectively filter out misjudged conflict events and improve the accuracy of recognition. The key to this step is to reasonably set the second preset threshold. The selection of the threshold needs to consider multiple factors, such as the traffic flow at the intersection, road conditions, and the type of traffic participants. For example, for intersections with heavy traffic, a smaller threshold may need to be set to improve safety; for intersections with less traffic, the threshold can be appropriately relaxed to reduce false alarms.

[0109] More specifically, the potential traffic safety analysis method of the embodiment of the present application provides multi-dimensional data support for identifying suspected longitudinal conflict events through the comprehensive utilization of motion trajectory information, motion parameter information and size information, and not only considers the motion state of traffic participants, but also considers their physical characteristics, so as to more comprehensively capture potential conflict situations; secondly, when predicting collision time information, the potential traffic safety analysis method of the embodiment of the present application makes full use of motion parameter information and size information, and not only considers the current motion state of traffic participants, but also considers the impact of their physical size on potential collisions; finally, potential longitudinal conflict events are screened by comparison with preset thresholds. This step provides an adjustable balancing mechanism for the entire identification process, so as to flexibly control the sensitivity of conflict identification, adapt to the characteristics of different types of traffic participants, and provide more accurate conflict warnings, thereby effectively improving the level of traffic safety.

[0110] In some preferred implementations, in order to further simplify the process of determining suspected longitudinal conflict events, step S321 preferably includes:

[0111] The overlap of the motion areas of any two traffic participants in each frame is analyzed by traversing the motion trajectory information and motion parameter information to extract the traffic participant combination whose overlap meets the longitudinal relationship; then, based on the motion parameter information of the traffic participant combination and the preset judgment table (such as Table 1), it is judged whether the traffic participant combination has a suspected longitudinal conflict event:

[0112] Table 1 Judgment table of suspected vertical conflict events

[0113]

[0114] Where a is the acceleration, a F is the acceleration of the following traffic participant, a L is the acceleration of the preceding traffic participant, V F is the speed of the following traffic participant, V Lis the speed of the preceding traffic participant, P represents the possibility of a traffic conflict, C represents the occurrence of a traffic conflict, and I represents the impossibility of a traffic conflict.

[0115] Specifically, the combination of traffic participants whose motion areas overlap in a longitudinal relationship represents that the motion paths of the two traffic participants are basically the same. Then, combined with Table 1, a comprehensive analysis can be made based on the speeds and accelerations of the two traffic participants to determine whether a traffic conflict is likely to occur between the two traffic participants, and whether the two traffic participants in the corresponding frame will trigger a suspected longitudinal conflict event, wherein the suspected longitudinal conflict event corresponds to the relationship between P and C in Table 1.

[0116] More specifically, the potential traffic safety analysis method of the embodiment of the present application can quickly determine suspected longitudinal conflict events based on the comparison of motion parameter information in Table 1, and can efficiently complete the extraction of suspected longitudinal conflict events.

[0117] In some embodiments, the collision time information can be calculated and obtained using a TTC (Time to Collision) algorithm, but this algorithm only considers the relationship between the distance and speed between vehicles to predict the collision time information, which is quite different from actual operation. Therefore, the potential traffic safety analysis method of an embodiment of the present application introduces acceleration information on this basis to improve the TTC algorithm to achieve accurate prediction of collision time information.

[0118] In some preferred embodiments, the motion parameter information includes velocity information and acceleration information;

[0119] Step S322 includes:

[0120] S3221, calculating collision travel information when the rear traffic participant moves to overlap with the initial position of the front traffic participant according to the speed information, acceleration information and size information of the traffic participant corresponding to the suspected longitudinal collision event;

[0121] S3222. Calculate and obtain collision time information based on the collision travel information and the speed information and acceleration information of the corresponding traffic participant.

[0122] Specifically, the potential traffic safety analysis method of the embodiment of the present application introduces speed information and acceleration information as motion parameters, providing a more comprehensive and accurate data basis for subsequent collision time prediction. This method takes into account the dynamic change characteristics of traffic participants and can more accurately reflect the actual traffic situation. Specifically, it can be achieved by the following steps:

[0123] First, the initial position of the traffic participant in front and the initial position, velocity and acceleration information of the traffic participant behind are obtained. Secondly, the relative position relationship between the two when the collision occurs is determined based on the size information of the traffic participants. Then, the kinematic formula is used to calculate the distance required for the traffic participant behind to catch up with the initial position of the traffic participant in front, combined with the velocity information and acceleration information.

[0124] After obtaining the collision travel information, the potential traffic safety analysis method of the embodiment of the present application further uses this information in combination with the speed and acceleration information of the traffic participants to calculate the collision time. This method can more accurately predict the time point when a collision may occur by considering the dynamic change characteristics of the traffic participants.

[0125] More specifically, based on the foregoing, it can be seen that the movement paths of the two traffic participants corresponding to the suspected longitudinal conflict event are basically the same. Therefore, step S3221 can combine the path distance, speed information and acceleration information of the two traffic participants in the corresponding frame images to quickly predict the moving distance of the rear traffic participant when their positions coincide (since no rear-end collision actually occurred, it can be simplified to position overlap for conversion and judgment) as the collision travel information.

[0126] More specifically, since the traffic participants have a certain volume, in order to obtain more accurate collision stroke information, step S3221 introduces size information to calculate the collision stroke information. In the embodiment of the present application, the size information in step S3221 includes the length information of the traffic participants in their moving direction.

[0127] More specifically, in the embodiment of the present application, step S3221 preferably calculates and obtains the collision stroke information by iterative calculation, as follows:

[0128] like Figure 3 As shown, the time t of the rear traffic participant after the corresponding frame is defined as n-1 to n The moving distance between moments is S (n-1)n , then it satisfies:

[0129] (1)

[0130] Among them, (X 2（tn） , Y 2（tn） , Z 2（tn） ) is the number of traffic participants behind at t n The three-dimensional coordinates at the time, (X 2（tn-1） , Y 2（tn-1） , Z 2（tn-1） ) is the number of traffic participants behind at t n-1 The three-dimensional coordinates of the moment;

[0131] Then, after the following traffic participant is in the corresponding frame, n The difference between the time and the initial position of the preceding traffic participant is △S （n-1）n ,satisfy:

[0132] (2)

[0133] Among them, (X 1（t0） , Y 1（t0） , Z 1（t0） ) is the three-dimensional coordinate of the preceding traffic participant at time t0 (i.e., the time corresponding to the frame image); therefore, it can be determined whether the following traffic participant has moved to a state where it overlaps with the initial position of the preceding traffic participant by determining whether the following formula is true:

[0134] (3)

[0135] Among them, L1 is the length information of the traffic participant in front, and L2 is the length information of the traffic participant in the back.

[0136] On this basis, step S3221 is transformed into iterative acquisition of the time t that meets formula (3) n , at the time t that meets formula (3) n After that, the collision travel information is recorded as D, which is calculated based on the following formula:

[0137] (4)

[0138] Specifically, the above calculation method introduces the length information in the size information representing the occupied space of the traffic participants, which can more accurately analyze and obtain the collision travel information so as to make the subsequent collision time information calculation more accurate, so as to accurately screen out potential longitudinal conflict events.

[0139] In some preferred embodiments, the collision time information is calculated as follows:

[0140] , △a≠0 (5)

[0141] , △a≠0 (6)

[0142] Wherein, △V is the speed difference between the rear traffic participant and the front traffic participant, △a is the acceleration difference between the rear traffic participant and the front traffic participant, and D is the collision travel information;

[0143] When both t1 and t2 are positive values, the collision time is the smaller value of t1 and t2; when only one of t1 and t2 is positive, the collision time is the positive value of t1 and t2.

[0144] Specifically, the method for calculating collision time information in the potential traffic safety analysis method of the embodiment of the present application takes into account the relative motion characteristics between traffic participants, and can more accurately predict the time when a collision may occur, thereby helping to timely identify potential traffic safety risks. Its advantage lies in that it takes into account speed differences and acceleration differences, and can reflect the complex relative motion state between traffic participants. By introducing collision stroke information, spatial factors are taken into consideration. In addition, two possible collision times are provided, and reasonable selection rules are given to improve the accuracy of the prediction. Its calculation method is clear and unambiguous, and easy to implement in an actual system.

[0145] More specifically, ΔV refers to the speed difference between the rear traffic participant and the front traffic participant in the frame image, which can be calculated based on ΔV=V F -V L Calculated, V F is the speed information of the following traffic participants, V L is the speed information of the preceding traffic participant; △a refers to the acceleration difference between the following traffic participant and the preceding traffic participant in the frame image, which can be calculated based on △a=a F -a L Calculated, a F is the acceleration information of the following traffic participant, a L It is the acceleration information of the preceding traffic participant.

[0146] More specifically, the collision time information is defined as MTTC, and MTTC satisfies:

[0147] If t1>0 and t2>0, and t1≥t2, then MTTC=t2;

[0148] If t1>0 and t2≤0, then MTTC=t1;

[0149] If t1≤0 and t2>0, and t1≥t2, then MTTC=t2.

[0150] In addition, it should be noted that if △a=0, then MTTC=D / △V, △V≠0.

[0151] The potential traffic safety analysis method of the embodiment of the present application provides a selection rule for the collision time. When both t1 and t2 are positive values, the smaller value is selected as the collision time; when only one is a positive value, the positive value is selected as the collision time. This selection method ensures the rationality of the calculation result; this calculation method can more accurately reflect the actual traffic situation and take into account the complex relative motion state between traffic participants. Compared with simple distance or speed threshold judgment, this method provides a more detailed and dynamic collision risk assessment.

[0152] More specifically, in the embodiment of the present application, the second preset threshold is preferably set to 1 or 4 seconds, and by comparing the magnitude relationship between MTTC and the second preset threshold, it can be determined whether the suspected longitudinal conflict event is a potential longitudinal conflict event.

[0153] In some preferred embodiments, step S32 further includes:

[0154] S324: Calculate collision index information according to the collision time information and the motion parameter information of the corresponding traffic participant.

[0155] Specifically, collision time information provides the key basis for judging the time when a conflict occurs, motion parameter information reflects the motion state of traffic participants, and collision index information represents a quantitative indicator of the degree of conflict. These features work together to achieve a quantitative assessment of longitudinal conflict events.

[0156] More specifically, the potential traffic safety analysis method of the embodiment of the present application can more objectively evaluate the danger level of longitudinal collision events by introducing collision index information as an auxiliary judgment criterion. By comprehensively considering time information and motion parameters, a scientific collision assessment mechanism is established, which effectively improves the accuracy and reliability of traffic safety analysis.

[0157] More specifically, in the embodiment of the present application, the collision index information is defined as CI, which satisfies:

[0158] (7)

[0159] More specifically, different potential vertical conflict events may have the same MTTC value. When this happens, the CI indicator is used to assist in indicating the severity of the conflict. When the MTTC values ​​are the same, the larger the CI value, the more severe the conflict.

[0160] In some more preferred embodiments, step S32 further comprises the steps of:

[0161] S325 : Compare the collision index information with a third preset threshold value to further screen potential longitudinal conflict events.

[0162] In some preferred embodiments, step S2 comprises:

[0163] S21, acquiring basic trajectory information and size information associated with the time series of each traffic participant according to the traffic participant trajectory data set;

[0164] S22, smoothing the basic trajectory information to obtain motion trajectory information;

[0165] S23, obtaining motion parameter information according to the smoothed motion trajectory information.

[0166] Specifically, the smoothing process can use algorithms such as Kalman filtering, Gaussian filtering or sliding average to eliminate noise interference in the trajectory data. In the embodiment of the present application, it is preferred to use a univariate interpolation method to interpolate the basic trajectory information to obtain smooth motion trajectory information of traffic participants.

[0167] More specifically, compared to directly using the original trajectory data, the potential traffic safety analysis method of the embodiment of the present application significantly improves the data quality through smoothing processing, providing more accurate input information for traffic conflict analysis. This processing method is applicable to all types of traffic participants and has strong reliability and practicality.

[0168] In some preferred embodiments, step S4 comprises:

[0169] S41, establishing a first distribution regularity diagram of traffic conflict frequency with respect to time of potential conflict event information of a target intersection based on the pairing relationship between conflict time information and traffic signal timing information;

[0170] S42: establishing a second distribution regularity diagram of traffic conflict frequency with respect to position of potential conflict event information at a target intersection based on the pairing relationship between the conflict position information and the road feature information.

[0171] Specifically, the potential traffic safety analysis method of the embodiment of the present application establishes a two-dimensional distribution law diagram by correlating traffic conflicts with time and space information, thereby achieving a comprehensive analysis of the temporal and spatial distribution law of traffic conflict frequencies at intersections. This analysis method can help identify potential safety risks at intersections at different time periods and locations, and provide data support for traffic safety management.

[0172] First distribution pattern chart: This chart can be in many forms, such as a bar chart, heat map or line chart. The horizontal axis can represent time (such as time within a signal cycle or different times of the day), and the vertical axis represents the frequency of conflicts. Different types of conflicts can be represented by different colors or patterns.

[0173] Second distribution pattern map: This chart can be in the form of a heat map or a three-dimensional surface map. The intersection area can be divided into a grid, and the conflict frequency at different locations can be represented by color depth or height.

[0174] More specifically, Figure 4It is a representation of the first distribution law diagram, wherein the color gradient bar at the top of the diagram represents the signal switching timing of the traffic signal timing information, the vertical axis in the diagram represents the number of occurrences of potential conflict event information, and the two curves represent the number of occurrences (i.e., traffic conflict frequency) of potential conflict event information generated by setting different benchmarks (first preset threshold and second preset threshold of different sizes) in the corresponding traffic signal cycle (90s), wherein the two vertical axis lines in the diagram correspond to the number of occurrences of potential conflict event information of different benchmarks, such as the second preset threshold corresponding to the left vertical axis line is 4 seconds, and the second preset threshold corresponding to the right vertical axis line is 1 second.

[0175] More specifically, Figure 5 It is a representation of the second distribution law diagram, which is drawn in the form of a heat map, where the straight lines in the figure represent the road distribution corresponding to the road feature information, and the color depth of the heat block in the figure represents the frequency of occurrence of the potential conflict event information generated at the corresponding position in the corresponding traffic signal cycle (90s) (i.e., traffic conflict frequency).

[0176] More specifically, compared to the traditional analysis method based only on accident data, the potential traffic safety analysis method of the embodiment of the present application has the following advantages: 1. Comprehensiveness: Through the analysis of the two dimensions of time and space, a comprehensive view of the frequency distribution of traffic conflicts at intersections is provided. 2. Preventiveness: Based on the analysis of potential conflict events, potential risks can be identified before an accident occurs, which has a stronger preventive effect. 3. Refinement: Through the correlation analysis with signal timing and road characteristics, a more refined and specific distribution law can be obtained. 4. Visualization: Through two distribution law diagrams, the complex spatiotemporal distribution law is intuitively presented, which is easy to understand and make decisions.

[0177] In some preferred embodiments, step S4 further includes:

[0178] S43, generating a spatiotemporal distribution regularity diagram of traffic conflict frequency of potential conflict event information at a target intersection by combining the first distribution regularity diagram and the second distribution regularity diagram.

[0179] Specifically, the potential traffic safety analysis method of the embodiment of the present application can obtain a graph that comprehensively reflects the distribution pattern of traffic conflict frequency in both time and space dimensions by integrating the first distribution pattern graph (time dimension) and the second distribution pattern graph (spatial dimension). Such a time-space distribution pattern graph can more intuitively and comprehensively display the traffic safety conditions at the intersection, and help traffic management departments to formulate more accurate traffic safety improvement measures.

[0180] More specifically, the spatiotemporal distribution pattern diagram can be drawn in the form of a heat map. For example, the originally complex terrain is transformed into a linear relationship according to the distance relationship design of the road characteristics as the vertical coordinate, and then the traffic signal cycle of the traffic signal timing information is set as the horizontal coordinate. Then, according to the frequency of traffic conflicts, a heat block set in the corresponding two-dimensional coordinate is established to obtain a visual spatiotemporal distribution pattern diagram.

[0181] More specifically, Figure 6 It is a representation of the spatiotemporal distribution law diagram, which establishes the vertical coordinate based on the minimum distance between the conflict location point and the stop line in the intersection. The color gradient bars at the top of the figure represent the signal switching form of the traffic signal timing information, that is, the traffic signal cycle is set as the horizontal coordinate, and then the vertical color gradient bars on the right are used to explain the traffic conflict frequency corresponding to the color of the thermal block.

[0182] Compared with the prior art, the potential traffic safety analysis method of the embodiment of the present application overcomes the limitation of considering only the time or space dimension by comprehensively considering the two dimensions of time and space, and can more comprehensively reflect the traffic safety status of the intersection. The innovation of this method lies in the organic combination of time and space information, providing a more comprehensive and efficient potential traffic safety analysis tool.

[0183] Second, please refer to Figure 7 Some embodiments of the present application also provide a potential traffic safety analysis device for analyzing traffic conflict rules, the device comprising:

[0184] A data acquisition module 201 is used to acquire traffic signal timing information, road feature information, and a traffic participant trajectory data set for at least one traffic signal cycle at a target intersection;

[0185] A feature extraction module 202 is used to obtain the time-series motion trajectory information, motion parameter information and size information of each traffic participant according to the traffic participant trajectory data set;

[0186] The conflict analysis module 203 is used to extract all potential conflict event information according to the motion trajectory information, motion parameter information and size information, and the potential conflict event information includes: conflict position information and conflict time information;

[0187] The conflict reconstruction module 204 is used to respectively match the road feature information and the traffic signal timing information based on the conflict position information and the conflict time information to obtain the spatiotemporal distribution law of the traffic conflict frequency at the target intersection.

[0188] The potential traffic safety analysis device of the embodiment of the present application overcomes the limitations of traditional accident data analysis devices and can evaluate the safety status of intersections in real time and dynamically. By analyzing the actual behavior of traffic participants rather than relying solely on accident data that has occurred, the device can identify potential safety risks earlier, thereby improving the foresight and preventive nature of safety analysis. At the same time, the device can intuitively display the distribution patterns of traffic conflicts in time and space by combining road characteristics and traffic signal timing information, and can more comprehensively consider various factors affecting intersection safety, thereby improving the accuracy and reliability of the analysis results and providing a quantitative basis for improving traffic safety.

[0189] In some preferred implementations, the potential traffic safety analysis device of the embodiments of the present application is used to execute the potential traffic safety analysis method provided in the first aspect above.

[0190] Third, please refer to Figure 8 Some embodiments of the present application also provide a structural diagram of an electronic device. The present application provides an electronic device, including: a processor 301 and a memory 302. The processor 301 and the memory 302 are interconnected and communicate with each other through a communication bus 303 and / or other forms of connection mechanisms (not marked). The memory 302 stores computer-readable instructions executable by the processor 301. When the electronic device is running, the processor 301 executes the computer-readable instructions to execute the method in any optional implementation of the above embodiments.

[0191] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method in any optional implementation of the above embodiment is executed. The computer-readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, disk or optical disk.

[0192] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0193] In addition, the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0194] Furthermore, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0195] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0196] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A potential traffic safety analysis method for analyzing traffic conflict patterns, characterized in that: The method comprises the following steps: S1, obtaining traffic signal timing information, road feature information and a traffic participant trajectory data set about at least one traffic signal cycle at a target intersection; S2. Acquire time-series motion trajectory information, motion parameter information, and size information of each traffic participant according to the traffic participant trajectory data set; S3, extracting all potential conflict event information according to the motion trajectory information, motion parameter information and size information, wherein the potential conflict event information includes: conflict position information and conflict time information; S4, based on the conflict position information and the conflict time information, respectively pairing the road feature information and the traffic signal timing information to obtain the spatiotemporal distribution law of the traffic conflict frequency at the target intersection; The types of the potential conflict event information are divided into potential horizontal conflict events and potential vertical conflict events. Step S3 includes: S30, determining the type of the traffic participant based on the size information of the traffic participant, and shielding the conflict analysis between pedestrians according to the type; S31, obtaining all suspected lateral conflict events in each frame and predicting the corresponding conflict time difference according to the motion trajectory information, motion parameter information and size information, and comparing the conflict time difference with a first preset threshold to filter the suspected lateral conflict events to obtain the potential lateral conflict events; S32, obtaining all suspected longitudinal conflict events in each frame and predicting the corresponding collision time information according to the motion trajectory information, motion parameter information and size information, and filtering the suspected longitudinal conflict events by comparing the collision time information with a second preset threshold value to obtain the potential longitudinal conflict event; Step S32 includes: S321, acquiring all suspected vertical conflict events in each frame according to the motion trajectory information, motion parameter information and size information; S322, predicting corresponding collision time information according to motion parameter information and size information of the traffic participant corresponding to the suspected longitudinal collision event; S323, comparing the collision time information with a second preset threshold value to filter suspected longitudinal conflict events to obtain potential longitudinal conflict events; S324, calculating collision index information according to the collision time information and the motion parameter information of the corresponding traffic participant; S325, comparing the collision index information and the third preset threshold value to further screen potential longitudinal conflict events; The collision index information is CI, which satisfies: ； Among them, △V is the speed difference between the rear traffic participant and the front traffic participant, △a is the acceleration difference between the rear traffic participant and the front traffic participant, MTTC is the collision time information, and a F is the acceleration of the following traffic participant, a L is the acceleration of the preceding traffic participant, V F is the speed of the following traffic participant, V L is the speed of the preceding traffic participant; Step S4 includes: S41, establishing a first distribution regularity diagram of traffic conflict frequency with respect to time of the potential conflict event information of the target intersection based on the pairing relationship between the conflict time information and the traffic signal timing information; S42, establishing a second distribution regularity diagram of traffic conflict frequency with respect to position of the potential conflict event information of the target intersection based on the pairing relationship between the conflict position information and the road feature information; S43. A spatiotemporal distribution pattern diagram of the traffic conflict frequency of the potential conflict event information of the target intersection is generated by integrating the first distribution pattern diagram and the second distribution pattern diagram. The process of drawing the spatiotemporal distribution pattern diagram includes: converting the terrain into a linear relationship as the vertical coordinate according to the distance relationship of the road feature information, and then setting the traffic signal period of the traffic signal timing information as the horizontal coordinate, and then establishing a thermal block set in the corresponding two-dimensional coordinate according to the traffic conflict frequency of the potential conflict event information.

2. The potential traffic safety analysis method according to claim 1, characterized in that: Step S31 includes: S311, obtaining conflict position information of all suspected lateral conflict events in each frame according to the motion trajectory information and size information; S312, predicting the conflict time difference of the traffic participant arriving at the conflict position information according to the conflict position information and the motion trajectory information and motion parameter information of the corresponding traffic participant; S313: Compare the conflict time difference with a first preset threshold value to filter out the suspected lateral conflict events and obtain the potential lateral conflict events.

3. The potential traffic safety analysis method according to claim 1, characterized in that: The motion parameter information includes speed information and acceleration information; Step S322 includes: S3221, calculating collision travel information when the rear traffic participant moves to overlap with the initial position of the front traffic participant according to the speed information, acceleration information and size information of the traffic participant corresponding to the suspected longitudinal collision event; S3222. Calculate and obtain the collision time information according to the collision travel information and the speed information and acceleration information of the corresponding traffic participant.

4. A potential traffic safety analysis device for analyzing traffic conflict patterns, characterized in that: The device comprises: A data acquisition module, used to acquire traffic signal timing information, road feature information, and a data set of traffic participant trajectories for at least one traffic signal cycle at a target intersection; A feature extraction module, used for obtaining the time-series motion trajectory information, motion parameter information and size information of each traffic participant according to the traffic participant trajectory data set; A conflict analysis module, used to extract all potential conflict event information according to the motion trajectory information, motion parameter information and size information, wherein the potential conflict event information includes: conflict position information and conflict time information; A conflict reconstruction module, used to respectively pair the road feature information and the traffic signal timing information based on the conflict position information and the conflict time information to obtain the spatiotemporal distribution law of the traffic conflict frequency of the target intersection; The types of the potential conflict event information are divided into potential lateral conflict events and potential longitudinal conflict events. The step of extracting all potential conflict event information according to the motion trajectory information, motion parameter information and size information includes: S30, determining the type of the traffic participant based on the size information of the traffic participant, and shielding the conflict analysis between pedestrians according to the type; S31, obtaining all suspected lateral conflict events in each frame and predicting the corresponding conflict time difference according to the motion trajectory information, motion parameter information and size information, and comparing the conflict time difference with a first preset threshold to filter the suspected lateral conflict events to obtain the potential lateral conflict events; S32, obtaining all suspected longitudinal conflict events in each frame and predicting the corresponding collision time information according to the motion trajectory information, motion parameter information and size information, and filtering the suspected longitudinal conflict events by comparing the collision time information with a second preset threshold value to obtain the potential longitudinal conflict event; Step S32 includes: S321, acquiring all suspected vertical conflict events in each frame according to the motion trajectory information, motion parameter information and size information; S322, predicting corresponding collision time information according to motion parameter information and size information of the traffic participant corresponding to the suspected longitudinal collision event; S323, comparing the collision time information with a second preset threshold value to filter suspected longitudinal conflict events to obtain potential longitudinal conflict events; S324, calculating collision index information according to the collision time information and the motion parameter information of the corresponding traffic participant; S325, comparing the collision index information and the third preset threshold value to further screen potential longitudinal conflict events; The collision index information is CI, which satisfies: ； Among them, △V is the speed difference between the rear traffic participant and the front traffic participant, △a is the acceleration difference between the rear traffic participant and the front traffic participant, MTTC is the collision time information, and a F is the acceleration of the following traffic participant, a L is the acceleration of the preceding traffic participant, V F is the speed of the following traffic participant, V L is the speed of the preceding traffic participant; The step of respectively pairing the road feature information and the traffic signal timing information based on the conflict position information and the conflict time information to obtain the spatiotemporal distribution law of the traffic conflict frequency at the target intersection comprises: S41, establishing a first distribution regularity diagram of traffic conflict frequency with respect to time of the potential conflict event information of the target intersection based on the pairing relationship between the conflict time information and the traffic signal timing information; S42, establishing a second distribution regularity diagram of traffic conflict frequency with respect to position of the potential conflict event information of the target intersection based on the pairing relationship between the conflict position information and the road feature information; S43. A spatiotemporal distribution pattern diagram of the traffic conflict frequency of the potential conflict event information of the target intersection is generated by integrating the first distribution pattern diagram and the second distribution pattern diagram. The process of drawing the spatiotemporal distribution pattern diagram includes: converting the terrain into a linear relationship as the vertical coordinate according to the distance relationship of the road feature information, and then setting the traffic signal period of the traffic signal timing information as the horizontal coordinate, and then establishing a thermal block set in the corresponding two-dimensional coordinate according to the traffic conflict frequency of the potential conflict event information.

5. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the steps in the method according to any one of claims 1 to 3 are executed.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 3 are executed.

Images