Pedestrian-vehicle traffic early warning system applied to traffic safety

Abstract

The invention discloses a human-vehicle traffic early warning system applied to traffic safety, which relates to the technical field of traffic safety and comprises a data acquisition unit, a data processing unit, an early warning decision unit, a user interaction unit, a safety and compliance unit and a system management and maintenance unit. The data processing unit is used for receiving the trajectory data and processing the trajectory data; the early warning decision-making unit is used for analyzing the processed trajectory data by applying a safe distance technology and a traffic conflict model so as to evaluate and predict traffic risks and generate an early warning decision; and the user interaction unit is used for displaying the generated early warning decision through a visual interface and sending out a corresponding early warning signal. Trajectory data of vehicles and pedestrians can be collected through the data acquisition unit and transmitted to the data processing unit for processing, a solid foundation is provided for subsequent early warning decisions, and timeliness and accuracy of early warning information are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of traffic safety, and in particular to a pedestrian and vehicle traffic warning system applied to traffic safety. Background Art

[0002] With the rapid development of the transportation industry, road traffic has increased significantly, which has led to an increase in the rate of traffic accidents. Faced with this challenge, many companies have developed navigation warning functions to help reduce traffic accidents. For example, a new function based on the high-precision positioning service of the Beidou satellite navigation system combined with a self-developed spatiotemporal perception model can predict the intersection of vehicles at non-light-controlled intersections and remind drivers of potential safety risks through the navigation interface and voice. The current navigation warning system already has a variety of functions, such as vehicle collision warning, lane departure reminder, and front vehicle start reminder. However, although these warning functions have achieved certain results in the application between vehicles, there are still obvious deficiencies in the warning between people, non-motor vehicles and vehicles. Most of the existing navigation warning systems are only applicable to the interaction between vehicles, and lack effective support for the multi-interaction warning between people, non-motor vehicles and vehicles, resulting in a serious lack of traffic conflict objects and an inability to fully cover all potential traffic safety risks.

[0003] Research on traffic prediction between vehicles has received widespread attention, especially multimodal trajectory prediction methods for autonomous driving based on deep learning, and prediction methods based on Transformer models, which have shown high prediction accuracy and generalization ability. However, when achieving multimodal high-precision prediction, these methods still face challenges such as information loss, interaction, and difficulty in modeling uncertainty. In addition, although some studies have proposed innovative methods such as the concept of a pedestrian safety system that combines on-site processing and centralized processing, and a predicted collision risk (PCR) assessment model based on a long short-term memory (LSTM) network, these studies still have difficulties in dealing with vehicle-pedestrian interactions and lack key data support such as signal phase information.

[0004] Research on traffic prediction between vehicles has also made significant progress. For example, the short-term traffic flow prediction method combining the K nearest neighbor (KNN) algorithm and the long short-term memory (LSTM) network model, as well as the vehicle trajectory prediction model based on the graph convolutional neural network (GCN), have improved the prediction accuracy to a certain extent. However, these research models still have some limitations, such as insufficient generalization ability, overfitting risk, and insufficient extraction of vehicle spatiotemporal interaction features, which makes it difficult to obtain accurate and effective vehicle trajectory prediction results in practical applications, and cannot fully meet the needs of human-vehicle traffic warning systems.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related technology, the present invention proposes a pedestrian and vehicle traffic warning system applied to traffic safety to overcome the above-mentioned technical problems existing in the existing related technology.

[0007] To this end, the specific technical solution adopted by the present invention is as follows:

[0008] A pedestrian and vehicle traffic warning system for traffic safety, the system comprising: a data acquisition unit, a data processing unit, a warning decision unit, a user interaction unit, a safety and compliance unit and a system management and maintenance unit;

[0009] A data acquisition unit, used to collect trajectory data of vehicles and pedestrians, and transmit the trajectory data to a data processing unit;

[0010] A data processing unit, used for receiving and processing trajectory data;

[0011] The early warning decision unit is used to analyze the processed trajectory data using safe distance technology and traffic conflict models to assess and predict traffic risks and generate early warning decisions;

[0012] A user interaction unit is used to display the generated warning decision through a visual interface and issue a corresponding warning signal;

[0013] The security and compliance unit is used to encrypt the trajectory data and perform security checks and vulnerability scans;

[0014] The system management and maintenance unit is used to monitor the operating status of the early warning system in real time, collect feedback data, and optimize the early warning system based on the feedback data.

[0015] Furthermore, the data acquisition unit includes: a data collection module and a data transmission module;

[0016] A data collection module, used to collect trajectory data of vehicles and pedestrians using sensors, the trajectory data including location information and road structure data;

[0017] The data transmission module is used to transmit the collected trajectory data to the data processing unit using the network library technology.

[0018] Furthermore, the data processing unit includes: a data receiving module, a data preprocessing module, an interactive sample generating module and a model input module;

[0019] A data receiving module, used to receive and store trajectory data to construct an interactive natural driving dataset;

[0020] The data preprocessing module is used to apply the sliding window technology to divide the trajectory data in the interactive natural driving dataset into several trajectory samples, and perform coordinate transformation processing on each trajectory sample;

[0021] An interaction sample generation module is used to combine the converted interaction vehicle trajectory data with the trajectory data of the predicted target to obtain interaction frames, and summarize the obtained interaction frames to obtain interaction samples;

[0022] The model input module is used to add the prediction start symbol in the decoded trajectory as the input of the decoder, train the pre-built trajectory prediction model, and use the prediction start symbol to verify and test the trained trajectory prediction model.

[0023] Furthermore, the data preprocessing module includes: a trajectory segmentation submodule, a sample segmentation submodule and a coordinate conversion submodule;

[0024] The trajectory segmentation module is used to sample and intercept trajectory samples based on the trajectory data in the interactive natural driving dataset using the sliding window technology;

[0025] The sample segmentation submodule is used to segment the captured trajectory samples into historical trajectories, decoded trajectories and target trajectories according to the preset segmentation criteria;

[0026] The coordinate conversion submodule is used to convert the coordinates of the remaining frames in the interactive natural driving dataset into relative coordinates relative to the reference frame based on the segmented trajectory samples and taking the coordinates of the reference frame as the origin coordinates.

[0027] Furthermore, based on the trajectory data in the interactive natural driving dataset, the sliding window technology is used to sample and intercept trajectory samples, including:

[0028] Based on the trajectory data in the interactive natural driving dataset, determine the parameters of the sliding window, including the window length and step size;

[0029] Using sliding window techniques, trajectory segments are extracted from interactive naturalistic driving datasets;

[0030] Verify the trajectory segment. If the trajectory segment in the window is smaller than the preset sliding window length, calculate the number of missing frames and set the missing number of frames as the new sliding step length.

[0031] The trajectory data within the window is intercepted as the trajectory sample, and the window is moved according to the new sliding step size until all the trajectory data are traversed.

[0032] Furthermore, the early warning decision unit includes: a risk definition module, an early warning generation module and an early warning information transmission module;

[0033] The risk definition module uses safe distance technology and pre-built traffic conflict models to conduct in-depth analysis of the processed trajectory data, assess the risk of vehicle-pedestrian collision, and set the risk level based on the degree of overlap of the collision area;

[0034] The warning generation module is used to generate corresponding warning information according to the set risk level;

[0035] The warning information transmission module is used to send the warning information to the user interaction unit in the form of an alarm prompt.

[0036] Further, the risk definition module includes: an intersection detection submodule, a collision prediction submodule and a risk level definition submodule;

[0037] The intersection detection submodule is used to set the safety zone using the safety distance technology and detect whether the safety zones between pedestrians and drivers intersect;

[0038] A collision prediction submodule is used to predict whether a collision will occur between a pedestrian and a driver using a pre-built traffic conflict model if an intersection of safety zones is detected;

[0039] The risk level definition submodule is used to evaluate the collision risk according to the prediction results and the degree of overlap between the vehicle and pedestrian collision areas, and set the corresponding risk level based on the evaluation results.

[0040] Furthermore, the safety distance technology is used to set the safety zone and detect whether the safety zones of pedestrians and drivers intersect, including:

[0041] Collect the movement information of pedestrians and drivers and calculate the corresponding initial safety distance. The movement information includes position data, speed, and acceleration.

[0042] Based on the initial safety distance, the safety distance is calculated by combining the driver's characteristic parameters and the characteristics of the automatic emergency braking process. The safety distance includes the warning safety distance and the driving safety distance.

[0043] Based on the calculated safety distance, the corresponding safety zones for pedestrians and drivers are set, and whether the safety zones between pedestrians and drivers intersect is detected;

[0044] If there is no intersection, the movement continues; otherwise, collision prediction is performed.

[0045] Furthermore, the formula for the warning safety distance is:

[0046]

[0047] Where, d w Indicates the warning safety distance value;

[0048] τ1 represents the driver’s reaction time;

[0049] t3 represents the time for canceling the pedal idle travel;

[0050] t4 represents the braking force boost time;

[0051] v1 represents the initial speed of the vehicle;

[0052] a represents the maximum deceleration;

[0053] d0 represents the minimum safety distance.

[0054] Furthermore, the formula for safe driving distance is:

[0055]

[0056] In the formula, S1 represents the driving safety distance value;

[0057] t1 represents the driver’s reaction time;

[0058] t2 represents the time it takes for the foot to move from the accelerator to the brake pedal;

[0059] t3 represents the time for canceling the pedal idle travel;

[0060] t4 represents the braking force boost time;

[0061] v1 represents the initial speed of the vehicle;

[0062] a represents the maximum deceleration;

[0063] d0 represents the minimum safety distance.

[0064] The beneficial effects of the present invention are:

[0065] 1. The present invention can collect trajectory data of vehicles and pedestrians through the data acquisition unit and transmit it to the data processing unit for processing, which provides a solid foundation for subsequent warning decisions and ensures the timeliness and accuracy of warning information.

[0066] 2. The present invention analyzes the processed trajectory data through the safety distance technology and traffic conflict model of the early warning decision unit, realizes accurate prediction of traffic risks between vehicles, non-motor vehicles and pedestrians, generates reliable early warning decisions, effectively reduces conflicts between people and vehicles, and improves road safety.

[0067] 3. The present invention displays the warning information generated by the warning decision unit in an intuitive manner through the user interaction unit, and simultaneously sends out a clear warning signal, ensuring that the user can quickly and accurately understand the warning content, and then take appropriate avoidance actions in time, thereby effectively preventing the occurrence of traffic accidents and ensuring road safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0069] Figure 1 is a principle block diagram of a pedestrian and vehicle traffic warning system applied to traffic safety according to an embodiment of the present invention;

[0070] Figure 2 is a schematic diagram of the overall structure of a pedestrian and vehicle traffic warning system applied to traffic safety according to an embodiment of the present invention;

[0071] Figure 3 is a flow chart of a human-vehicle traffic warning application program in a human-vehicle traffic warning system for traffic safety according to an embodiment of the present invention;

[0072] Figure 4 is a technical architecture diagram of a pedestrian and vehicle traffic warning application program in a pedestrian and vehicle traffic warning system for traffic safety according to an embodiment of the present invention;

[0073] Figure 5 is a workflow diagram of a pedestrian and vehicle traffic warning system applied to traffic safety according to an embodiment of the present invention;

[0074] Figure 6 is a schematic diagram of a pedestrian in a pedestrian and vehicle traffic warning system applied to traffic safety according to an embodiment of the present invention;

[0075] Figure 7 is a schematic diagram of a vehicle in a pedestrian and vehicle traffic warning system for traffic safety according to an embodiment of the present invention;

[0076] Figure 8 is a schematic diagram of the relationship between road surface, vehicle speed and braking distance in a pedestrian and vehicle traffic warning system applied to traffic safety according to an embodiment of the present invention;

[0077] Fig. 9 The figure is a schematic diagram of deceleration of a pedestrian and vehicle traffic warning system applied to traffic safety according to an embodiment of the present invention.

[0078] In the figure:

[0079] 1. Data collection unit; 2. Data processing unit; 3. Early warning decision unit; 4. User interaction unit; 5. Security and compliance unit; 6. System management and maintenance unit. DETAILED DESCRIPTION

[0080] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments. They can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention.

[0081] According to an embodiment of the present invention, a pedestrian and vehicle traffic warning system for traffic safety is provided.

[0082] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1-Figure 2 As shown, according to an embodiment of the present invention, the pedestrian and vehicle traffic warning system applied to traffic safety comprises: a data collection unit 1, a data processing unit 2, a warning decision unit 3, a user interaction unit 4, a safety and compliance unit 5 and a system management and maintenance unit 6;

[0083] The data collection unit 1 is used to collect trajectory data of vehicles and pedestrians and transmit the trajectory data to the data processing unit 2.

[0084] Specifically, the data acquisition unit 1 includes: a data collection module and a data transmission module;

[0085] A data collection module, used to collect trajectory data of vehicles and pedestrians using sensors, the trajectory data including location information and road structure data;

[0086] The data transmission module is used to transmit the collected trajectory data to the data processing unit 2 using the network library technology.

[0087] The data processing unit 2 is used to receive the trajectory data and process the trajectory data.

[0088] Specifically, the data processing unit 2 includes: a data receiving module, a data preprocessing module, an interactive sample generation module and a model input module;

[0089] The data receiving module is used to receive and store trajectory data to construct an interactive natural driving dataset.

[0090] It should be noted that the Interactive Natural Driving Dataset (INTERACTION Natural Driving Dataset) uses drone and high-altitude camera image acquisition systems with trajectory tracking and data denoising methods, lists the data types such as trajectories and road structures contained in the dataset, explains the properties and characteristics of the data in the dataset, and performs data preprocessing such as trajectory segmentation, trajectory data smoothing, interactive vehicle selection, and matching of interactive vehicles with predicted targets.

[0091] The data preprocessing module is used to apply the sliding window technology to divide the trajectory data in the interactive natural driving dataset into several trajectory samples, and perform coordinate transformation processing on each trajectory sample.

[0092] Specifically, the data preprocessing module includes: a trajectory segmentation submodule, a sample segmentation submodule and a coordinate conversion submodule;

[0093] The trajectory segmentation submodule is used to sample and intercept trajectory samples based on the trajectory data in the interactive natural driving dataset using the sliding window technology.

[0094] Specifically, based on the trajectory data in the interactive natural driving dataset, the sliding window technology is used to sample and intercept trajectory samples, including:

[0095] Based on the trajectory data in the interactive natural driving dataset, determine the parameters of the sliding window, including the window length and step size;

[0096] Using sliding window techniques, trajectory segments are extracted from interactive naturalistic driving datasets;

[0097] Verify the trajectory segment. If the trajectory segment in the window is smaller than the preset sliding window length, calculate the number of missing frames and set the missing number of frames as the new sliding step length.

[0098] The trajectory data within the window is intercepted as the trajectory sample, and the window is moved according to the new sliding step size until all the trajectory data are traversed.

[0099] The sample segmentation submodule is used to segment the captured trajectory samples into historical trajectories, decoded trajectories and target trajectories according to the preset segmentation criteria.

[0100] The coordinate conversion submodule is used to convert the coordinates of the remaining frames in the interactive natural driving dataset into relative coordinates relative to the reference frame based on the segmented trajectory samples and taking the coordinates of the reference frame as the origin coordinates.

[0101] The interaction sample generation module is used to combine the converted interaction vehicle trajectory data with the trajectory data of the predicted target to obtain interaction frames, and summarize the obtained interaction frames to obtain interaction samples.

[0102] The model input module is used to add the prediction start symbol in the decoded trajectory as the input of the decoder, train the pre-built trajectory prediction model, and use the prediction start symbol to verify and test the trained trajectory prediction model.

[0103] It should be added that the trajectory prediction model uses Beidou-GPS positioning and machine learning technology to predict human-vehicle conflicts.

[0104] Splitting the trajectory data means that in each trajectory, a sliding window is used to calculate and intercept the trajectory sample. The sliding window length is 60 frames (i.e. 6 seconds) and the sliding step is 30 frames. If the trajectory data in the window is less than 60 frames after the window moves, the number of missing frames is calculated and the number of frames is set as the sliding step. After obtaining the trajectory sample, the sample needs to be split into three parts: historical trajectory (first 30 frames), decoded trajectory (31st frame-59th frame), and target trajectory (31st-60th frame); the historical trajectory is used to input the temporal feature encoder, the decoded trajectory is used to input the decoder, and the target trajectory is used for error calculation during the trajectory prediction model training process. The vehicle coordinates of the reference frame are changed to the origin coordinates (0, 0), and the absolute position coordinates of the vehicle in other frames in the entire trajectory sample are processed as relative coordinates relative to the reference frame coordinates. The coordinate values ​​contained in the historical trajectory are negative, and the coordinate values ​​contained in the decoded trajectory and the target trajectory are positive.

[0105] The trajectory information of the interactive vehicle and the predicted target are combined into an interactive frame and a flag is set, where the predicted target is recorded as 1 and the interactive vehicle is recorded as 0; finally, 30 interactive frames are combined into an interactive sample, and the timestamp of each frame in the interactive sample needs to correspond to the timestamp of each frame in the trajectory sample of the predicted target. For the input data of the trajectory prediction model decoder, since the input data of the decoder in the training phase is different from that in the verification and testing phase, two different lengths of data are required. In the trajectory prediction model training phase, the prediction start symbol and the future real trajectory of the vehicle (i.e., the decoded trajectory) need to be input as the prediction start prompt and "prediction reference", so the prediction start symbol needs to be added before the first frame of the decoded trajectory, and (0, 0, 0) is used as the prediction start symbol. In the trajectory prediction model verification and testing phase, the decoder only needs to input the prediction start symbol.

[0106] The early warning decision unit 3 is used to analyze the processed trajectory data using the safety distance technology and the traffic conflict model to evaluate and predict traffic risks and generate early warning decisions.

[0107] Specifically, the early warning decision unit 3 includes: a risk definition module, an early warning generation module and an early warning information transmission module;

[0108] The risk definition module is used to use safe distance technology and pre-built traffic conflict models to conduct in-depth analysis of the processed trajectory data, evaluate the risk of vehicle-pedestrian collision, and set the risk level based on the degree of overlap of the collision area.

[0109] Specifically, the risk definition module includes: an intersection detection submodule, a collision prediction submodule, and a risk level definition submodule;

[0110] The intersection detection submodule is used to set the safety area using the safety distance technology and detect whether there is an intersection between the safety areas of pedestrians and drivers.

[0111] Specifically, the safe distance technology is used to set the safe area and detect whether the safe areas between pedestrians and drivers intersect, including:

[0112] Collect the motion information of pedestrians and drivers and calculate the corresponding initial safety distance. The motion information includes position data, speed, and acceleration.

[0113] Based on the initial safety distance, the safety distance is calculated in combination with the driver's characteristic parameters and the characteristics of the automatic emergency braking process. The safety distance includes the warning safety distance and the driving safety distance.

[0114] Specifically, the formula for the warning safety distance is:

[0115]

[0116] Where, d w Indicates the warning safety distance value;

[0117] τ1 represents the driver’s reaction time;

[0118] t3 represents the time for canceling the pedal idle travel;

[0119] t4 represents the braking force boost time;

[0120] v1 represents the initial speed of the vehicle;

[0121] a represents the maximum deceleration;

[0122] d0 represents the minimum safety distance.

[0123] Specifically, the formula for safe driving distance is:

[0124]

[0125] In the formula, S1 represents the driving safety distance value;

[0126] t1 represents the driver’s reaction time;

[0127] t2 represents the time it takes for the foot to move from the accelerator to the brake pedal;

[0128] t3 represents the time for canceling the pedal idle travel;

[0129] t4 represents the braking force boost time;

[0130] v1 represents the initial speed of the vehicle;

[0131] a represents the maximum deceleration;

[0132] d0 represents the minimum safety distance.

[0133] Based on the calculated safety distance, the corresponding safety areas for pedestrians and drivers are set, and it is detected whether there is an intersection between the safety areas of pedestrians and drivers.

[0134] If there is no intersection, the movement continues; otherwise, a collision prediction is performed.

[0135] The collision prediction submodule is used to predict whether a collision will occur between a pedestrian and a driver using a pre-built traffic conflict model if an intersection of the safety zones is detected.

[0136] The risk level definition submodule is used to evaluate the collision risk according to the prediction results and the degree of overlap between the vehicle and pedestrian collision areas, and set the corresponding risk level based on the evaluation results.

[0137] It should be noted that the state frame of the prediction target at time t in the past is defined as X t =(x t ,y t , flag), where x t ,y t To predict the target position coordinates at time t, To predict the heading angle of the target at time t, flag represents the vehicle role (1: predicted target, 0: interactive vehicle). Define the historical trajectory sequence of the predicted target with a duration of H+1 as X=(X -H , X -H+1 ), …, X t , …, X 0 ), where t 0 , X 0 is the reference frame (the last frame of the historical trajectory sequence). Secondly, define the multi-vehicle mixed interaction frame C composed of the interactive vehicle and the predicted target at time t t =C t =(C1 t , C2 t , …, C n t , X t ), where Cn t =(x t ,y t , flag) represents the state of the interactive vehicle n at time t, x t ,y t is the relative position coordinate of the interactive vehicle relative to the predicted target at time t, is the heading angle of the interactive vehicle at time t. Define a multi-vehicle mixed interactive frame sequence C with a duration of H+1 = (C H , C H+1 , C t , C0), where t0, C0 are reference frames, X and C are both time series and each frame in the sequence is kept corresponding in time. The trajectory coordinate point of the predicted target at time t in the future is defined as Define the future trajectory sequence of the predicted target with a duration of T as Y = (Y 1 , Y 2 , …, Y t , …, Y T ).

[0138] The Traffic Conflict Model (PCR Model) defines the risk level based on the degree of overlap between the vehicle and pedestrian collision areas (PCRA), as follows:

[0139] Danger: If the PCRA of the vehicle and pedestrian overlap within 1 second;

[0140] Warning: If the PCRA of a vehicle and a pedestrian overlap within 2 seconds;

[0141] Relatively safe: If the vehicle and pedestrian overlap in PCRA after 3 seconds;

[0142] Normal: There is no overlap in PCRA.

[0143] In addition, if Figure 6-Figure 9 As shown, vulnerable road users (VRUs) are at high risk in traffic accidents, so systems need to be designed to prevent accidents and respond proactively.

[0144] After receiving the information, pedestrians need about 1.2 seconds to react and make a decision. After receiving the warning information, pedestrians may change their original walking route or behavior pattern to avoid potential danger.

[0145] Drivers need two steps before making decisions: cognitive reaction time (PRT), including perception, cognition, action planning and initial preparation time before starting vehicle control actions; maneuvering time (MT): mainly depends on external factors such as vehicle performance, road surface, tires, road geometry, etc. Among them, the inertia of the vehicle during driving; the driver's reaction to the signal takes a certain amount of time; the limitation of the vehicle's braking capacity.

[0146] The driver characteristic parameter value range is selected as [0, 1] and integrated with the automatic emergency braking process (AEB braking process) to obtain the driver characteristic parameter. The relationship between the driver reaction time t1 and the minimum safety distance d0 is:

[0147]

[0148] Warning safety distance d w The formula is:

[0149]

[0150] Where, d w Indicates the warning safety distance value;

[0151] τ1 represents the driver’s reaction time;

[0152] t3 represents the time for canceling the pedal idle travel;

[0153] t4 represents the braking force boost time;

[0154] v1 represents the initial speed of the vehicle;

[0155] a represents the maximum deceleration;

[0156] d0 represents the minimum safety distance.

[0157] Assume the vehicle's initial speed v1 (m / s) and maximum deceleration a (m / s 2 ), the braking distance S after speed v1 is:

[0158]

[0159] In the formula, S represents the safe distance value for braking start;

[0160] t1 represents the driver’s reaction time;

[0161] t2 represents the time it takes for the foot to move from the accelerator to the brake pedal;

[0162] t3 represents the time for canceling the pedal idle travel;

[0163] t4 represents the braking force boost time;

[0164] v1 represents the initial speed of the vehicle;

[0165] a represents the maximum deceleration.

[0166] When the car brakes, it needs to keep a minimum safety distance d0 from the obstacle in front. Then the sum of the travel distance and the minimum safety distance when the car completes braking is S1:

[0167]

[0168] In the formula, S1 represents the driving safety distance value;

[0169] t1 represents the driver’s reaction time;

[0170] t2 represents the time it takes for the foot to move from the accelerator to the brake pedal;

[0171] t3 represents the time for canceling the pedal idle travel;

[0172] t4 represents the braking force boost time;

[0173] v1 represents the initial speed of the vehicle;

[0174] a represents the maximum deceleration;

[0175] d0 represents the minimum safety distance.

[0176] In addition, the average adhesion coefficients on various road surfaces are shown in Table 1:

[0177] Table 1 Average adhesion coefficient on various road surfaces

[0178] Serial number pavement Peak adhesion coefficient Sliding adhesion coefficient 1 Asphalt or concrete (dry) 0.8～0.9 0.75 2 Asphalt (wet) 0.5～0.7 0.45～0.6 3 Concrete (wet) 0.8 0.7 4 gravel 0.6 0.55 5 Dirt road (dry) 0.68 0.65 6 Dirt road (wet) 0.55 0.4～0.5 7 Snow (compacted) 0.2 0.15 8 ice 0.1 0.07

[0179] The warning generation module is used to generate corresponding warning information according to the set risk level.

[0180] The warning information transmission module is used to send the warning information to the user interaction unit in the form of an alarm prompt.

[0181] The user interaction unit 4 is used to display the generated warning decision through a visual interface and issue a corresponding warning signal.

[0182] The security and compliance unit 5 is used to encrypt the trajectory data and perform security checks and vulnerability scans.

[0183] The system management and maintenance unit 6 is used to monitor the operating status of the early warning system in real time, collect feedback data, and optimize the early warning system based on the feedback data.

[0184] In a specific embodiment, Figure 3As shown in the figure, a smart watch is used to obtain the geographical location of pedestrians and upload it to the mobile phone for analysis, and the data is further aggregated to the server for storage. It can be mainly divided into two ports: the client, the application (APP) is installed on the user's mobile phone and smart watch, responsible for presenting the user interface, processing user input and displaying warning information; the server is responsible for processing complex calculations, storing large amounts of data, maintaining user account information, etc.

[0185] Among them, the application is a traffic warning APP for people and vehicles, such as Figure 4 As shown, Android studio is used as the environment for application development and debugging. The framework structure of the program is written in XML language in the activity main interface (activity_main.xml) file in the environment, and the design view is used to assist in modifying the structure. XML language is used to write styles in the style settings (styles.xml) file, and Kotlin in the activity class file is used for logical interaction between pages (such as page jumps, function definition, etc.), and the user's geographic location information is obtained by calling the location interface (Location interface).

[0186] like Figure 5 As shown in the figure, the watch uses the Location interface technology in Android studio to obtain the geographical location, speed, and acceleration of pedestrians and drivers and transmit them to the mobile phone. The mobile phone sends the user's data to the backend server through the network library technology, and uses the server's storage function to increase data capacity and realize data sharing to meet the data requirements of subsequent models. At the same time, the mobile phone will provide a map page in the APP and display the location of all users. Users can observe the surrounding traffic conditions in real time. Using the safety distance technology in traffic engineering, each user's safety circle can be made. The mobile phone can determine the risk by detecting whether there is an intersection between the ranges of users. If there is an intersection, the established traffic conflict model can be used to further predict whether there will be a conflict between pedestrians and drivers. Based on the wearable interface technology that comes with Android studio, the watch can receive the warning requirements sent by the mobile phone through the APP, and use vibration or sound warnings to remind users that an accident is about to happen to avoid potential dangers.

[0187] Through smart devices such as smart watches, users can receive real-time warning information from APP or traffic management systems. This instant notification method greatly reduces the risk of traffic accidents caused by distraction from checking mobile phones. At the same time, smart watches can also monitor the driver's vital signs, such as heart rate, stress level, etc. Once an abnormality is found, it will immediately issue a warning to prompt the driver to stop and rest in time, thereby effectively avoiding safety hazards such as fatigue driving.

[0188] In summary, with the aid of the above technical solution of the present invention, the trajectory data of vehicles and pedestrians can be collected by the data acquisition unit 1 and transmitted to the data processing unit 2 for processing, which provides a solid foundation for subsequent early warning decisions and ensures the timeliness and accuracy of early warning information. Through the safety distance technology and traffic conflict model of the early warning decision unit 3, the processed trajectory data is analyzed to achieve accurate prediction of traffic risks between vehicles, non-motor vehicles and pedestrians, generate reliable early warning decisions, effectively reduce conflicts between people and vehicles, and improve road safety. The early warning information generated by the early warning decision unit 3 is displayed in an intuitive manner through the user interaction unit 4, and a clear early warning signal is issued simultaneously, ensuring that the user can quickly and accurately understand the content of the early warning, and then take appropriate avoidance actions in time, thereby effectively preventing the occurrence of traffic accidents and ensuring road safety.

[0189] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A pedestrian and vehicle traffic warning system applied to traffic safety, characterized in that: The system includes: a data acquisition unit, a data processing unit, an early warning decision unit, a user interaction unit, a safety and compliance unit, and a system management and maintenance unit; The data acquisition unit is used to collect trajectory data of vehicles and pedestrians and transmit the trajectory data to the data processing unit; The data processing unit is used to receive the trajectory data and process the trajectory data; The warning decision unit is used to analyze the processed trajectory data using the safety distance technology and the traffic conflict model to evaluate and predict traffic risks and generate warning decisions; The user interaction unit is used to display the generated warning decision through a visual interface and issue a corresponding warning signal; The security and compliance unit is used to encrypt the trajectory data and perform security checks and vulnerability scans; The system management and maintenance unit is used to monitor the operating status of the early warning system in real time, collect feedback data, and optimize the early warning system based on the feedback data.

2. A traffic warning system for people and vehicles used for traffic safety according to claim 1, characterized in that: The data acquisition unit includes: a data collection module and a data transmission module; The data collection module is used to collect trajectory data of vehicles and pedestrians using sensors, wherein the trajectory data includes location information and road structure data; The data transmission module is used to transmit the collected trajectory data to the data processing unit using the network library technology.

3. The pedestrian and vehicle traffic warning system for traffic safety according to claim 1 is characterized in that: The data processing unit includes: a data receiving module, a data preprocessing module, an interactive sample generating module and a model input module; The data receiving module is used to receive and store trajectory data to construct an interactive natural driving data set; The data preprocessing module is used to apply the sliding window technology to divide the trajectory data in the interactive natural driving data set into a plurality of trajectory samples, and perform coordinate transformation processing on each trajectory sample; The interaction sample generation module is used to combine the converted interaction vehicle trajectory data with the trajectory data of the predicted target to obtain interaction frames, and summarize the obtained interaction frames to obtain interaction samples; The model input module is used to add a prediction start symbol to the decoded trajectory as the input of the decoder, train the pre-built trajectory prediction model, and verify and test the trained trajectory prediction model using the prediction start symbol.

4. A traffic warning system for people and vehicles used for traffic safety according to claim 3, characterized in that: The data preprocessing module includes: a trajectory segmentation submodule, a sample segmentation submodule and a coordinate conversion submodule; The trajectory segmentation submodule is used to sample and intercept trajectory samples using a sliding window technology according to the trajectory data in the interactive natural driving dataset; The sample segmentation submodule is used to segment the intercepted trajectory samples into historical trajectories, decoded trajectories and target trajectories according to a preset segmentation standard; The coordinate conversion submodule is used to convert the coordinates of the remaining frames in the interactive natural driving data set into relative coordinates relative to the reference frame based on the segmented trajectory samples and taking the coordinates of the reference frame as the origin coordinates.

5. A traffic warning system for people and vehicles used for traffic safety according to claim 4, characterized in that: The method of sampling and intercepting trajectory samples by using sliding window technology according to the trajectory data in the interactive natural driving dataset includes: Determine the parameters of the sliding window based on the trajectory data in the interactive natural driving dataset, wherein the parameters include the window length and the step length; Using sliding window techniques, trajectory segments are extracted from interactive naturalistic driving datasets; Verify the trajectory segment. If the trajectory segment in the window is smaller than the preset sliding window length, calculate the number of missing frames and set the missing number of frames as the new sliding step length. The trajectory data within the window is intercepted as the trajectory sample, and the window is moved according to the new sliding step size until all the trajectory data are traversed.

6. The pedestrian and vehicle traffic warning system for traffic safety according to claim 1 is characterized in that: The warning decision unit includes: a risk definition module, a warning generation module and a warning information transmission module; The risk definition module is used to use the safety distance technology and the pre-built traffic conflict model to conduct in-depth analysis of the processed trajectory data, evaluate the risk of collision between vehicles and pedestrians, and set the risk level according to the degree of overlap of the collision area; The warning generation module is used to generate corresponding warning information according to the set risk level; The warning information transmission module is used to send the warning information to the user interaction unit in the form of an alarm prompt.

7. A traffic warning system for people and vehicles used for traffic safety according to claim 6, characterized in that: The risk definition module includes: an intersection detection submodule, a collision prediction submodule and a risk level definition submodule; The intersection detection submodule is used to set the safety area using the safety distance technology and detect whether there is an intersection between the safety areas of the pedestrian and the driver; The collision prediction submodule is used to predict whether a collision will occur between a pedestrian and a driver using a pre-built traffic conflict model if an intersection of safety areas is detected; The risk level definition submodule is used to evaluate the collision risk according to the prediction results and the degree of overlap between the vehicle and pedestrian collision areas, and set a corresponding risk level based on the evaluation results.

8. The pedestrian and vehicle traffic warning system for traffic safety according to claim 7 is characterized in that: The method of setting a safety zone by using the safety distance technology and detecting whether the safety zones between pedestrians and drivers intersect includes: Collecting movement information of pedestrians and drivers and calculating the corresponding initial safety distance, wherein the movement information includes position data, speed, and acceleration; Based on the initial safety distance, the safety distance is calculated in combination with the driver characteristic parameters and the characteristics of the automatic emergency braking process, wherein the safety distance includes the warning safety distance and the driving safety distance; Based on the calculated safety distance, the corresponding safety zones for pedestrians and drivers are set, and whether the safety zones between pedestrians and drivers intersect is detected; If there is no intersection, the movement continues; otherwise, collision prediction is performed.

9. A traffic warning system for people and vehicles used for traffic safety according to claim 8, characterized in that: The formula of the warning safety distance is: Where, d w Indicates the warning safety distance value; τ1 represents the driver’s reaction time; t3 represents the time for canceling the pedal idle travel; t4 represents the braking force boost time; v1 represents the initial speed of the vehicle; a represents the maximum deceleration; d0 represents the minimum safety distance.

10. The pedestrian and vehicle traffic warning system for traffic safety according to claim 8, characterized in that: The formula for the safe driving distance is: In the formula, S1 represents the driving safety distance value; t1 represents the driver’s reaction time; t2 represents the time it takes for the foot to move from the accelerator to the brake pedal; t3 represents the time for canceling the pedal idle travel; t4 represents the braking force boost time; v1 represents the initial speed of the vehicle; a represents the maximum deceleration; d0 represents the minimum safety distance.

Images