Test method and system for vehicle-human interaction safety based on real-time adaptive triggering of ttc

Abstract

This invention relates to the field of automotive safety testing technology, specifically to a human-vehicle interaction safety testing method and system based on real-time adaptive triggering (TTC). The method includes: constructing a multi-source heterogeneous data acquisition subsystem, including a sensing end, a soft-body vehicle target and a control end, and a global time synchronization module; pre-setting fixed kinematic parameters and physical safety boundaries for the soft-body vehicle target, calculating its complete kinematic curve from the starting point to the physical stopping point and a setpoint for the total motion duration, and pre-storing these values ​​in the control end; acquiring real-time motion state data of the two-wheeled vehicle during the test, calculating the collision time (TTC) to reach the preset collision reference point; when the TTC meets the triggering conditions related to the setpoint for the total motion duration, issuing a command to control the target to move according to the pre-stored curve; collecting and recording test data, aligning the time series based on the triggering time, and constructing a human-vehicle interaction database. This technical solution can improve the realism of the test scenario interaction and cover multi-dimensional riding behavior.

Description

Technical Field

[0001] This invention relates to the field of automotive safety testing technology, specifically to a human-vehicle interaction safety testing method and system based on TTC real-time adaptive triggering. Background Technology

[0002] With the development of intelligent connected vehicle technology, active safety features such as Automatic Emergency Braking (AEB) have become standard equipment in mass-produced vehicles, improving road traffic safety. However, in traffic scenarios involving vulnerable road users (VRUs) such as electric two-wheelers, such as crossing intersections, traveling straight in opposite directions, and suddenly exiting blind spots, the accident rate remains high. Evaluating active safety features in these scenarios is a crucial step in verifying the safety performance of intelligent vehicles.

[0003] Currently, the evaluation methods for active safety systems of intelligent vehicles are mainly based on industry regulations, using dummies, dummy vehicles, or software targets as test objects. The testing process is usually based on a predetermined static trajectory or uniform motion mode, which has limitations in human-vehicle interaction safety testing.

[0004] Specifically, existing technologies face the following technical challenges: First, the test scenarios lack interactivity and dynamic response capabilities. The software targets or dummies move according to a preset, fixed program, failing to simulate the rider's stress response when facing collision risks, such as emergency braking, sharp steering maneuvers, and other hazard avoidance actions. The programmed movement mode makes the test one-way, meaning the tested vehicle only responds to the target object; the target object cannot react in real-time to the tested vehicle's dynamics. This results in a difference between the interaction process between the test scenario and a real accident.

[0005] Secondly, the diversity and complexity of cycling behavior are difficult to reproduce. Cyclists on real roads exhibit a wide range of behavioral characteristics: conservative cyclists anticipate risks and proactively avoid them, opportunistic cyclists weave through traffic, and aggressive cyclists cross the road quickly; their movement patterns are highly variable and random. Existing technologies use standardized static or uniformly moving targets, which cannot cover the dynamic behavior of different groups of people. Therefore, test data cannot accurately reflect real-world accident scenarios, limiting the reference value of test results.

[0006] Third, there is a contradiction between the safety risks of real-vehicle testing and data acquisition. Using human riders and real cars to conduct high-risk tests increases the risk of collisions and poses personal safety hazards. This safety risk makes it difficult to obtain data on dangerous human-vehicle interactions, leaving algorithm optimization and verification without the support of real-world interaction data.

[0007] Furthermore, in existing software target object tests, the target object triggering mechanism relies on fixed distance or time thresholds, lacking adaptive triggering capabilities based on real-time collision time (TTC). When the motion state of a two-wheeled vehicle changes dynamically, fixed triggering logic cannot guarantee the accuracy and repeatability of the test scenario construction, and cannot meet the safety performance evaluation requirements under specific interaction sequences. Summary of the Invention

[0008] The purpose of this invention is to propose a human-vehicle interaction safety test method based on real-time adaptive triggering of TTC, which can retain the dynamic behavioral characteristics of real cyclists, improve the realism of the test scenario interaction, cover multi-dimensional cycling behavior, and ensure the personal safety of the test subjects.

[0009] To achieve the above objectives, in a first aspect, this invention proposes a human-vehicle interaction safety test method based on TTC real-time adaptive triggering, comprising: A multi-source heterogeneous data acquisition subsystem was built, including a sensing end for collecting rider and two-wheeled vehicle status data, a software-based vehicle target object and its control end as an interactive object, and a global time synchronization module. The fixed kinematic parameters and physical safety boundaries of the soft car target are preset, the complete kinematic curve of the soft car target from the starting point to the physical stopping point and the corresponding total motion duration are calculated, and the kinematic curve is pre-stored in the control terminal; During the experiment, the motion state data of the two-wheeled vehicle is acquired in real time, and the collision time (TTC) of the two-wheeled vehicle reaching the preset collision reference point is calculated in real time based on the motion state data. When the collision time (TTC) meets the triggering condition related to the set value of the total motion duration, a triggering command is sent to the control terminal to control the soft car target to perform motion according to the pre-stored kinematic curve. Collect and record multi-source heterogeneous data during the experiment, align the time series based on the trigger time, and construct a human-vehicle interaction database.

[0010] Beneficial effects of the basic scheme: This invention pre-calibrates the fixed kinematic parameters and physical safety boundaries of the soft-body vehicle target, inversely calculates and solidifies the kinematic curve and total motion duration of the target from start to stop, making the target's trajectory controllable and predictable. During the test, a real-time TTC collision time calculation mechanism triggers the target's motion based on the two-wheeled vehicle's driving state, physically and logically constraining the target's motion termination position, ensuring the target stops at a safe distance before the preset collision reference point, avoiding a physical collision between the two-wheeled vehicle and the target. This mechanism constructs a closed-loop safety protection system, requiring no manual emergency intervention, and supports high-frequency, high-intensity extreme hazard testing, improving test fault tolerance while reducing equipment wear and personnel risks.

[0011] This invention employs an adaptive TTC triggering logic based on the real-time motion state of a two-wheeled vehicle, replacing the traditional mechanical scenario simulation with fixed timing and speed. The system collects motion data such as the vehicle's speed, position, and attitude in real time, dynamically calculates the collision time, and triggers the target object's movement. This ensures that the intervention timing and driving rhythm of the soft-body vehicle target object closely match the actual driving state of the two-wheeled vehicle, reproducing the dynamic characteristics of typical high-risk traffic scenarios such as sudden pedestrian appearances, crossing blind spots at intersections, and close-range overtaking. This dynamic adaptation mechanism can induce the rider's emergency response, obtaining real-world data on the subject's hazard avoidance maneuvers and driving behavior. It solves the problems of rigid testing scenarios, stiff interaction, and distorted behavior in traditional testing methods, making the testing scenarios consistent with real road conditions. This provides scenario support for research on the human-machine interaction mechanism of two-wheeled vehicles and the performance verification of active safety algorithms.

[0012] In traditional human-vehicle interaction testing, heterogeneous data from multiple dimensions, such as inertial navigation dynamics data, multi-camera visual image data, and human physiological and electrical data, suffer from problems such as inconsistent acquisition clocks, timing misalignment, and insufficient synchronization accuracy. This invention establishes a global time synchronization module, using the GPS absolute timestamp as a unified reference clock to calibrate the time of multi-source data acquired by the sensing and control ends. The system uses the target trigger time as a benchmark to standardize and align the time series of structured motion data, unstructured video image data, and human physiological and electrical data throughout the entire test process, controlling the multi-source data synchronization error to ≤0.02s. This synchronization mechanism achieves time-series matching of human-vehicle motion states, scene changes, and driver behavioral and physiological characteristics, eliminating test errors caused by data timing deviations and providing a data foundation for data analysis, model training, and performance verification.

[0013] The human-vehicle interaction database, built on a standardized collection and alignment mechanism, is comprehensive in data dimensions, time-series matching, and real in behavior. It can support scientific research work such as the iteration of active safety technology for two-wheeled vehicles, research on driver behavior mechanisms, and optimization of intelligent risk avoidance algorithms.

[0014] This invention constructs an integrated multi-source heterogeneous data acquisition subsystem comprising a sensing and acquisition terminal, a target object control terminal, and a global time synchronization module. Employing a modular architecture and standardized functional interfaces, it achieves the collection and management of comprehensive data, including rider status, two-wheeled vehicle operating conditions, interactive target object movement, and human physiological factors. Pre-stored and fixed target object kinematic curves, standardized trigger logic, and unified data alignment rules reduce experimental deviations caused by differences in manual operation and equipment timing, achieving uniformity in testing procedures, scenario parameters, and data standards. This system adapts to the testing needs of different vehicle models, road conditions, and hazardous scenarios, supports batch repeatable and comparative tests, and can accumulate a standardized human-vehicle interaction test database, providing technical support for the establishment of active safety testing standards for two-wheeled vehicles and the implementation of technological achievements.

[0015] As a feasible preferred embodiment, the fixed kinematic parameters and physical safety boundaries of the soft vehicle target are preset, including: Establish a local coordinate system for the test field, and define the geometric intersection point of the trajectory line of the soft car target and the preset cross trajectory line of the two-wheeled vehicle as the collision reference point; The physical stopping point of the soft car target is set at a fixed physical safety distance in front of the collision reference point; Based on the set target driving speed and motion limit parameters, the total motion time required for the soft car target to come to a complete stop at the physical stopping point is calculated, and the corresponding velocity-time function and acceleration-time function are generated as the kinematic curves.

[0016] As a feasible and preferred solution, the time to collision (TTC) for two-wheeled vehicles to reach a preset collision reference point is calculated in real time, including: Receive the real-time coordinates and speed of the two-wheeled vehicle; Calculate the real-time absolute distance from the current center of gravity of the two-wheeled vehicle to the collision reference point; The current collision time (TTC) is calculated based on the ratio of the real-time absolute distance to the real-time velocity.

[0017] As a feasible preferred option, the triggering condition is that the collision time TTC is less than or equal to the set value of the total motion duration.

[0018] As a feasible and preferred solution, dynamic randomization of free riding anti-predictive control measures: Multiple pass opportunities are set up in the test area. A randomization procedure is used to randomly select a portion of the multiple pass opportunities as the actual trigger count, and the remaining counts are used as blank test counts. In the actual number of triggers, a trigger command is issued when the trigger condition is met; in the blank test, no trigger command is issued even if the trigger condition is met, so that the soft car target remains stationary.

[0019] As a feasible and preferred solution, the multi-source heterogeneous data acquisition subsystem also includes a human factors data acquisition terminal for collecting the rider's physiological signals; the sensing terminal includes a combined navigation device mounted on the two-wheeled vehicle for outputting the real-time speed and absolute coordinates of the two-wheeled vehicle.

[0020] As a feasible and preferred approach, time series alignment based on the trigger time includes: The moment when the trigger command is issued is used as the global zero-point timestamp; A time window centered on the global zero-point timestamp is set, and the two-wheeled vehicle navigation data, target visual acquisition data, and rider physiological signal data within the time window are extracted and aligned.

[0021] As a feasible and preferred solution, the global time synchronization module uses the second pulse signal output by the GNSS receiver as the absolute time reference and synchronizes the internal clock source of each subsystem through the network time synchronization protocol, thereby controlling the absolute timestamp synchronization error of heterogeneous data within a preset threshold.

[0022] As a feasible preferred solution, the overall motion pattern of the soft car target object is regularized as a fixed curve process of stationary motion, uniform acceleration, uniform speed driving, uniform deceleration to stop. The soft car target is composed of a motion control platform carrying soft covering material or a standardized vehicle shape soft target. After receiving a trigger command, the control terminal starts from the current stationary state and strictly controls the platform to move according to the pre-stored kinematic curve until it stops at the physical stop point.

[0023] Secondly, the present invention also proposes a human-vehicle interaction safety test system based on TTC real-time adaptive triggering, which utilizes the aforementioned human-vehicle interaction safety test method based on TTC real-time adaptive triggering. Attached Figure Description

[0024] Figure 1 This is a logical diagram of a human-vehicle interaction safety test method based on TTC real-time adaptive triggering.

[0025] Figure 2 This is a schematic diagram of the interactive experimental system architecture and communication topology.

[0026] Figure 3 A schematic diagram of fixed kinematic parameter curves for a car target object.

[0027] Figure 4 This is a schematic diagram of the layout and triggering geometry of the interactive test site.

[0028] Figure 5 This is a schematic diagram of the absolute time alignment of multi-source heterogeneous data. Detailed Implementation

[0029] To make the technical solution and advantages of this application clearer, the technical solution of the present invention will be further described in detail below with reference to the accompanying drawings. It is understood that the specific embodiments described herein are only some embodiments of the present invention, and are only used to explain this application, not to limit it. It should be noted that the technical features or combinations of technical features described in the following embodiments should not be considered isolated; they can be combined with each other to achieve better technical effects. The same reference numerals appearing in the accompanying drawings of the following embodiments represent the same features or components, and can be applied to different embodiments.

[0030] Furthermore, unless otherwise defined, the technical or scientific terms used in the description of this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0031] The present invention will now be described in further detail with reference to the accompanying drawings.

[0032] Reference Figure 1 This disclosure provides a human-vehicle interaction safety test method based on TTC real-time adaptive triggering, including the following steps.

[0033] Step S1, construct a multi-source heterogeneous data acquisition subsystem based on absolute time synchronization: Deploy a multi-source sensing network within the closed test area. In one embodiment, the closed test area is a standard intersection within an automotive test track or an area capable of simulating traffic conflicts at an intersection. The overall system architecture refers to... Figure 2 As shown, it includes the following core subsystems and their deployment methods.

[0034] Step S1-1, the deployment method of the electric two-wheeled vehicle perception and data acquisition subsystem includes: Vehicle preparation: Select an electric two-wheeler with basic riding functions (such as an electric bicycle or electric moped). To ensure safety, the vehicle must undergo a safety inspection, and the braking and steering systems must be in good working order.

[0035] The integrated navigation system is installed by rigidly mounting a high-precision integrated navigation device (such as the i-TESTER AVE 2100 test system) at the vehicle's center of gravity (usually located on the frame near the seat). This system incorporates a GNSS receiver supporting multi-satellite, multi-frequency RTK (real-time dynamic differential) positioning and tightly coupled a high-precision microelectromechanical (MEMS) inertial measurement unit (IMU). Its core technical parameters must meet the following requirements: horizontal positioning accuracy better than 1.0 cm + 1 ppm, speed measurement accuracy better than 0.02 km / h, roll and pitch angle measurement accuracy better than 0.03° (in RTK fixed mode), and dual-antenna heading angle accuracy better than 0.1°. The device outputs the vehicle's absolute position, real-time speed, and dynamic acceleration, among other core status parameters, to the main control computer via a network or CAN bus at a frequency of at least 100 Hz.

[0036] The integrated navigation system is powered by either an onboard battery or a separate power pack. The device's data output interfaces (such as Ethernet or serial ports) are connected to the main control computer via cables.

[0037] Steps S1-2, the deployment methods of the soft vehicle target and its control subsystem include: The target carrier employs a motion control platform, such as a large flatbed trailer or a modified remote-controlled vehicle model (which must possess sufficient size, mass, and motion performance to simulate a real car), covered with soft-surface material or fitted with standardized soft targets in the shape of a vehicle. In this embodiment, it is referred to as a test flatbed moving system or a vehicle target.

[0038] The motion control platform includes a braking system, a low-level motion controller, and a high-precision positioning module.

[0039] The drive and braking system consists of a high-performance servo motor, a reducer, and a brake, and is capable of accurately executing preset acceleration, speed, and deceleration curves.

[0040] The underlying motion controller receives trajectory commands from the upper-level controller of the target object and performs closed-loop control to ensure that the actual motion is consistent with the commands.

[0041] The high-precision positioning module is equipped with an RTK-GNSS receiver, which provides real-time feedback of its own position for closed-loop calibration of motion control.

[0042] The upper-level controller of the target object preferably uses an industrial computer to deploy control software. It pre-stores the complete kinematic curves (velocity-time v(t) curve and acceleration-time a(t) curve) calculated offline according to step S2 from the starting point to the physical stopping point. It receives trigger commands from the main control computer via a wireless network. Once the trigger command is received, it immediately starts from the current stationary state (time t=0) and strictly controls the platform movement according to the pre-stored v(t) and a(t) curves until it comes to a complete stop at the set physical stopping point.

[0043] At the visual acquisition front end, a high frame rate industrial camera (e.g., more than 120 frames per second) is installed at the front of the target platform (simulating the bottom of a car's windshield), with the lens facing directly forward, to capture dynamic images of the two-wheeled vehicle crossing the path.

[0044] Steps S1-3, the multi-source heterogeneous data joint acquisition and synchronization subsystem, includes: From an aerial perspective, a multi-rotor drone is deployed approximately 30-50 meters above the conflict area at the intersection. Equipped with a high-definition, stabilized gimbal camera, the drone records a comprehensive overview of the entire intersection, the trajectories of two-wheeled vehicles, and the trajectories of other vehicles from a top-down angle. The drone transmits the video stream in real-time to a ground station via an image transmission system.

[0045] A global time synchronization center is established, connecting all heterogeneous subsystems. The center utilizes a 1PPS (pulse per second) hardware signal output from a local GNSS receiver as the absolute time reference. Each independent device synchronizes its internal hardware clock source via a network time synchronization protocol, strictly controlling the absolute timestamp synchronization error of heterogeneous data to within ≤0.02s.

[0046] The human data acquisition device consists of a wireless physiological signal acquisition system worn by each cyclist (subject) participating in the experiment, including: An electrocardiogram (ECG) or photoplethysmography (PPG) sensor, worn on the chest or wrist, acquires heart rate (HR) and heart rate variability (HRV) signals at a high frequency (e.g., 250 Hz).

[0047] Electrical conductance (EDA) sensors, typically consisting of two electrode patches, are attached to the palm or fingertips of the non-dominant hand of a subject to measure subtle changes in skin conductance levels.

[0048] The device should be able to send timestamped physiological data to the data aggregation node in real time via Bluetooth or a dedicated wireless network.

[0049] Step S2 involves presetting fixed kinematic parameters and physical safety boundaries for the target object, thereby eliminating the risk of human-vehicle physical collisions at the physical level. This embodiment employs feedforward fixed dynamic curve control. Step S2-1, establish the local coordinate system of the test field and define the collision reference point, including: Using high-precision RTK-GNSS equipment, differential point mapping was performed on the test area to establish a local Cartesian coordinate system for the test area. The geometric intersection of the center trajectory line of the soft-body vehicle target and the preset cross-trajectory line of the electric two-wheeled vehicle was defined as the collision reference point (P). col ); Step S2-2: Set the physical safety stop point and safety distance, and set the final physical stop point P of the soft vehicle target (test tablet moving system). STOP Set at a distance P col Fixed physical safety distance D in front safe (Strictly set at 3m). This distance is set to take into account both the control performance margin of the underlying motor of the target object and the rider's psychological safety threshold, to prevent secondary dangers such as crashes caused by extreme panic inducing the subject due to excessively close distance; Steps S2-3 involve regularizing the overall motion of the target vehicle into a fixed curve: stationary → uniform acceleration → uniform speed travel → uniform deceleration → stop, and setting its maximum acceleration, maximum deceleration, and acceleration limit (Jerk). Based on the set target speed v... target The total time T required for the target object to come to a complete stop from its initial stationary position to its physical stopping point can be accurately calculated offline. total .

[0050] The calculated velocity-time function v(t) (0≤t≤T_total) and acceleration-time function a(t) for the entire motion process are pre-stored into the target object's upper-level controller in the form of a discrete data point table or a parameterized formula. Figure 3 It displays a pre-defined, fixed kinematic curve.

[0051] In one embodiment, taking the scenario of a two-wheeled vehicle crossing at a distance of 60 km / h as an example, the target speed of the vehicle is set to 60 km / h, with a maximum acceleration of 2 m / s² and a maximum deceleration of 6 m / s². Calculations show that the vehicle needs to start 120 meters from the point of impact, accelerate, maintain a constant speed, and brake before precisely stopping 3 meters in front of the reference point. The total time T for this motion process is... total It is 12.78s.

[0052] Step S3, refer to Figure 4 Closed-loop adaptive TTC threshold trigger control. The subject drives a two-wheeled vehicle towards an intersection, and the system executes adaptive triggering logic, including: The two-wheeled vehicle-side integrated navigation system outputs real-time speed v(t) and absolute coordinates under high dynamic riding conditions.

[0053] The central control unit receives the real-time coordinates of the two-wheeled vehicle and calculates the distance from the vehicle's current center of gravity to the collision reference point P. col The real-time absolute distance d(t); The decision-making center calculates the current collision time in real time, using the following expression:

[0054] The system continuously compares Compared with the time setpoint T in step S2 total When the timing signal is detected to satisfy the condition for the first time... ≤T total At that moment, the main control unit immediately sends a trigger command to the software target controller, and the vehicle target responds instantly and executes a fixed motion curve.

[0055] In one embodiment, taking the scenario of "two-wheeled vehicle crossing at a distance of 60km / h" as an example, the system refreshes the TTC value of the two-wheeled vehicle at a frequency of 100Hz, and immediately issues a trigger command when TTC≤12.78s.

[0056] Step S4 involves dynamic randomization of free riding anti-expectation control to eliminate the adaptive defensive driving psychology of subjects in repeated tests.

[0057] Dynamic anti-predictive control logic was set up within the test area: volunteers were allowed to ride freely on a fixed route within the test area, and the main control center randomly triggered car targets according to the triggering logic throughout the entire ride.

[0058] This scheme does not completely limit the total number of test groups and the trigger ratio: the central control unit runs a randomization program. For example, the program is set to run a total of N_ tests within a complete test session (including multiple laps). otal = 10 “potentially triggered passes” (that is, each time a rider approaches the conflict zone, the system calculates TTC in the background to determine whether the triggering conditions are met).

[0059] Of these 10 opportunities, the system randomly selects M. trigger =7 times, when the TTC condition is met, the actual execution is triggered (the vehicle target is dispatched).

[0060] The remaining N total -M trigger Three times, even if the TTC conditions are met, the system does not send a trigger command, and the vehicle target remains stationary. These three times constitute the "blank test".

[0061] In traditional repetitive testing, subjects are prone to developing scene memory and psychological expectations, leading to defensive driving behavior. This causes the collected driving behavior data and avoidance response data to deviate from the actual road conditions. This embodiment of the present disclosure, by interspersing irregular, non-interactive routine riding in time, prevents subjects from establishing a conditioned reflex of "a car will appear every time I reach a certain location," disrupting their scene prediction rhythm and suppressing unnatural defensive driving behavior. Each time a subject approaches a conflict area, they cannot be certain whether a danger will occur, thus maintaining a relatively natural state of alertness and riding behavior. Only at the actual trigger moment can a more realistic stress response be induced. Step S5: Heterogeneous time series alignment and construction of a multi-dimensional high-fidelity interactive database. All data are aligned at the millisecond level based on the trigger moment.

[0062] After the trigger occurs, the data collection center uses TTC(t) = T total The time when the trigger command is issued is used as the global zero-point timestamp t0.

[0063] A time window centered at t0 is defined. Based on t0, data from each channel within this time window is automatically extracted and aligned, including Channel 1 (discrete point sequence output by integrated navigation), Channel 2 (conflict feature frames segmented from the camera in front of the target), and Channel 3 (heart rate changes and skin conductance mutation points extracted by the electrophysiological signal acquisition instrument within the corresponding time series). The aligned and standardized time series set is stored in a database to accurately identify dangerous scene elements and cluster cycling styles.

[0064] Figure 5 It demonstrates how data from multiple channels can achieve precise timing alignment along the vertical line t0.

[0065] This disclosure also provides a human-vehicle interaction safety test system based on TTC real-time adaptive triggering, which is applied to the above-mentioned human-vehicle interaction safety test method based on TTC real-time adaptive triggering.

[0066] The above content is merely an embodiment of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can improve and implement the present invention based on the guidance provided in this application and their own capabilities. Typical well-known structures or operating methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A human-vehicle interaction safety test method based on TTC real-time adaptive triggering, characterized in that, include: A multi-source heterogeneous data acquisition subsystem was built, including a sensing end for collecting rider and two-wheeled vehicle status data, a software-based vehicle target object and its control end as an interactive object, and a global time synchronization module. The fixed kinematic parameters and physical safety boundaries of the soft car target are preset, the complete kinematic curve of the soft car target from the starting point to the physical stopping point and the corresponding total motion duration are calculated, and the kinematic curve is pre-stored in the control terminal; During the experiment, the motion state data of the two-wheeled vehicle is acquired in real time, and the collision time (TTC) of the two-wheeled vehicle reaching the preset collision reference point is calculated in real time based on the motion state data. When the collision time (TTC) meets the triggering condition related to the set value of the total motion duration, a triggering command is sent to the control terminal to control the soft car target to perform motion according to the pre-stored kinematic curve. Collect and record multi-source heterogeneous data during the experiment, align the time series based on the trigger time, and construct a human-vehicle interaction database.

2. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, The fixed kinematic parameters and physical safety boundaries of the preset soft-body vehicle target object include: Establish a local coordinate system for the test field, and define the geometric intersection point of the trajectory line of the soft car target and the preset cross trajectory line of the two-wheeled vehicle as the collision reference point; The physical stopping point of the soft car target is set at a fixed physical safety distance in front of the collision reference point; Based on the set target driving speed and motion limit parameters, the total motion time required for the soft car target to come to a complete stop at the physical stopping point is calculated, and the corresponding velocity-time function and acceleration-time function are generated as the kinematic curves.

3. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, Real-time calculation of the time to collision (TTC) for two-wheeled vehicles to reach a preset collision reference point, including: Receive the real-time coordinates and speed of the two-wheeled vehicle; Calculate the real-time absolute distance from the current center of gravity of the two-wheeled vehicle to the collision reference point; The current collision time (TTC) is calculated based on the ratio of the real-time absolute distance to the real-time velocity.

4. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, The triggering condition is that the collision time TTC is less than or equal to the set value of the total motion duration.

5. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, Dynamic and randomized free riding anti-predictive control steps: Multiple pass opportunities are set up in the test area. A randomization procedure is used to randomly select a portion of the multiple pass opportunities as the actual trigger count, and the remaining counts are used as blank test counts. In the actual number of triggers, a trigger command is issued when the trigger condition is met; In the blank test, even if the triggering condition is met, no triggering command is issued, so that the soft car target remains stationary.

6. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, The multi-source heterogeneous data acquisition subsystem also includes a human factors data acquisition terminal for collecting the rider's physiological signals; the sensing terminal includes a combined navigation device mounted on the two-wheeled vehicle for outputting the real-time speed and absolute coordinates of the two-wheeled vehicle.

7. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, Time series alignment based on the trigger time includes: The moment when the trigger command is issued is used as the global zero-point timestamp; A time window centered on the global zero-point timestamp is set, and the two-wheeled vehicle navigation data, target visual acquisition data, and rider physiological signal data within the time window are extracted and aligned.

8. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, The global time synchronization module uses the second pulse signal output by the GNSS receiver as the absolute time reference and synchronizes the internal clock source of each subsystem through the network time synchronization protocol, so as to control the absolute timestamp synchronization error of heterogeneous data within a preset threshold.

9. The human-vehicle interaction safety test method based on TTC real-time adaptive triggering according to claim 1, characterized in that, The overall motion of the soft car target object is a fixed curve process of being stationary, accelerating uniformly, moving at a constant speed, decelerating uniformly to a stop; The soft car target is composed of a motion control platform carrying soft covering material or a standardized vehicle shape soft target. After receiving a trigger command, the control terminal starts from the current stationary state and strictly controls the platform to move according to the pre-stored kinematic curve until it stops at the physical stop point.

10. A human-vehicle interaction safety test system based on TTC real-time adaptive triggering, characterized in that, Applied to the human-vehicle interaction safety test method based on TTC real-time adaptive triggering as described in any one of claims 1-9.

Images