Vehicle risk collision time calculation method based on expected function safety

Abstract

The invention provides a vehicle risk collision time calculation method based on expected function safety, and relates to the technical field of automatic driving expected function safety. The method specifically comprises the following steps: constructing an automatic driving takeover simulation environment, setting a plurality of groups of different road scenes, takeover scenes and vehicle speeds in the automatic driving takeover simulation environment, and constructing a plurality of automatic driving system takeover scenes; performing an automatic driving takeover simulation experiment on each automatic driving system takeover scene to obtain takeover response time and physiological characteristic index data of a plurality of groups of drivers; and calculating weighted average takeover response time of the driver according to the takeover response time of the driver and the physiological characteristic index data, and further calculating the vehicle risk collision time by utilizing the weighted average takeover response time of the driver. The method can be used for ensuring the decision reasonability of the automatic driving system under the emergency condition.

Description

Technical Field

[0001] The present invention relates to the technical field of expected functional safety of autonomous driving, and in particular to a method for calculating vehicle risk collision time based on expected functional safety. Background Art

[0002] With the development and application of autonomous driving technology, L3 autonomous driving systems can independently complete driving tasks in specific scenarios, but in complex or beyond system capabilities, the driver needs to quickly take over vehicle control, which involves the issue of human-machine control switching. However, long-term autonomous driving mode will lead to a decrease in driver attention and alertness, affecting the driver's ability to respond in emergencies. Therefore, it is particularly important to study the psychological and behavioral characteristics of the driver's takeover process, as well as optimize takeover prompts and human-computer interaction technology. The study of the control switching problem also involves the decision-making logic and sensor data processing capabilities of the autonomous driving system, aiming to make accurate takeover judgments in complex traffic environments, and provide sufficient time after issuing a takeover request to help the driver quickly understand the current situation and take over safely. Therefore, the time node of the takeover request should not only take into account the impact of the external environment, but also the driver's reaction time to the takeover request. The study of L3 autonomous driving takeover issues is not only related to technical implementation, but also involves human-computer interaction, safety and system reliability, and is a key link in realizing the large-scale application of autonomous driving technology.

[0003] In the development of autonomous driving technology, how to make decisions and effectively avoid risks is a critical and complex issue. In this context, takeover and autonomous emergency braking (AEB) systems, as core components of the autonomous driving safety architecture, can effectively reduce the risk of collision. However, there are differences in the urgency and mechanism of these two safety measures, takeover and AEB. How to accurately determine the time window and ownership of control switching is a challenge that needs to be solved in the field of autonomous driving. Takeover means that when the autonomous driving system encounters complex traffic situations or technical failures, the system activates the corresponding mode through the human-computer interaction interface to remind the driver to quickly intervene and take over the control of the vehicle. However, in the takeover process, there are problems such as when to take over, how to effectively transfer control, and how to ensure that the driver can restore control of the vehicle in a timely and accurate manner, that is, how to effectively control the collision risk of the vehicle, thereby avoiding accidents and improving driving safety.

[0004] In previous studies, most of them use time to collision (TTC) to assess the risk of collision, and then help the autonomous driving system decide the next decision. TTC is a key concept in the field of autonomous driving, which is used to measure the time required for the vehicle to reach the vehicle in front or the obstacle at the current speed. It is usually calculated by dividing the distance between the vehicle and the obstacle by their relative speed. In addition, there is a method of using cognitive load to set the takeover reminder time. By monitoring the driver's physiological signals, such as heart rate, eye movement, etc., the reminder time is dynamically adjusted to provide personalized takeover prompts. Although it is highly complex, it can effectively avoid the risks caused by overload or distraction. At the same time, it combines environmental perception technology to rely on sensors such as lidar and cameras to obtain real-time environmental information, so as to dynamically adjust the takeover reminder time and judge the takeover demand based on road conditions, traffic conditions, weather and other factors.

[0005] For the method of using collision time to assess collision risk, this method mainly relies on the distance and relative speed between the vehicle and the obstacle to predict the collision time, without fully considering the driver's reaction time and cognitive differences. In actual scenarios, the driver may not be able to respond immediately due to inattention, fatigue or other external factors. Therefore, relying solely on TTC to determine the takeover reminder time may cause the system to underestimate the risk and delay the takeover timing. Especially in complex human-machine co-driving environments, TTC has low accuracy and cannot adapt to dynamically changing driving environments, such as other vehicles suddenly changing lanes and pedestrians suddenly appearing. In complex traffic scenarios, TTC's static algorithm cannot flexibly respond to rapidly changing situations, which may lead to incorrect takeover judgments. The method of using cognitive load method to set the takeover reminder time usually evaluates the driver's current cognitive state by real-time monitoring of physiological signals such as heart rate, brain waves, eye movements, etc. Although this method can provide personalized takeover reminders, the implementation of this method often requires complex sensor networks and data processing systems, resulting in high computational overhead and possible real-time performance. In addition, most of the devices used in this method are contact devices, which are difficult to use in real application scenarios. In addition, this method that relies on physiological data has extremely high requirements for the accuracy and robustness of sensors, which may lead to misjudgment due to equipment failure or data noise, and requires continuous collection of the driver's physiological data, which may involve privacy issues. Long-term monitoring of the driver's physiological signals may not only cause privacy concerns, but also increase the driver's psychological burden, thereby affecting the driving experience. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention proposes a vehicle risk collision time calculation method based on expected functional safety by designing a variant decision indicator of time to collision (TTC), aiming to ensure the rationality of the decision-making of the autonomous driving system in emergency situations.

[0007] The present invention proposes a method for calculating vehicle risk collision time based on expected functional safety, the method comprising the following steps:

[0008] Step 1: Build an autonomous driving takeover simulation environment, including: scenario simulation module, human-computer interaction system and data acquisition module;

[0009] Step 2: Set up several groups of different road scenarios, takeover scenarios, and vehicle speeds in the autonomous driving takeover simulation environment, and build several autonomous driving system takeover scenarios;

[0010] Step 3: Conduct an autonomous driving takeover simulation experiment for each autonomous driving system takeover scenario to obtain takeover reaction time and physiological characteristic index data of several groups of drivers;

[0011] Step 4: Calculate the weighted average takeover reaction time of the driver based on the takeover reaction time and physiological characteristic index data of the driver;

[0012] Step 5: Count the takeover reaction times of all drivers, and select the takeover reaction times that meet the preset ratio from the statistical results as redundant time, and calculate the vehicle risk collision time using the weighted average takeover reaction time and redundant time of the drivers;

[0013] The scenario simulation module described in step 1 is used to construct a vehicle dynamics model, takeover control logic and a preset test environment model that simulates the takeover scenario of the autonomous driving system;

[0014] The human-machine interaction system is used to manage the switching of the vehicle between the automatic driving state and the driver takeover state according to the takeover control logic. When the driver receives a takeover request from the vehicle, the driver manually controls the vehicle to switch from the automatic driving state to the driver takeover state; when the takeover is completed, the vehicle automatically switches from the driver takeover state to the automatic driving state;

[0015] The data acquisition module is used to collect the driver's physiological characteristic index data and the vehicle's motion state change data in the simulated automatic driving system takeover scenario;

[0016] The human-machine interaction system includes: vehicle status, interaction mode and driver part;

[0017] The vehicle state is used to simulate the real-time driving state of the vehicle during driving; the vehicle state includes: vehicle speed, vehicle position and steering wheel movement;

[0018] The interactive mode is used to simulate the prompting mode of the vehicle sending a takeover request to the driver; the interactive mode includes: auditory prompting, visual prompting and tactile prompting;

[0019] The driver part is used to simulate different driving states of the driver; the driver part includes: driver attributes and driving sub-tasks; wherein the driver attributes are used to simulate drivers with different driving conditions; the driver attributes include: the driver's age, gender and driving experience; the driving sub-tasks are used to simulate the driver's state when receiving a takeover request; the driving sub-tasks include: talking, sleeping, listening to music and playing games;

[0020] The data acquisition module includes: a physiological signal acquisition module and a vehicle motion acquisition module;

[0021] The physiological signal acquisition module is used to collect the driver's brain electrocardiogram signal data, respiratory rate data and skin temperature data in real time in a scenario where the driver simulates the automatic driving system taking over;

[0022] The vehicle motion acquisition module is used to collect the longitudinal speed change and lateral displacement of the vehicle during the continuous driving of the vehicle after the vehicle switches to the driver-takeover state; wherein the longitudinal speed change of the vehicle is used to record in real time the speed change of the vehicle during the continuous driving of the vehicle after the vehicle switches to the driver-takeover state; the lateral displacement of the vehicle is used to record in real time the lateral position displacement of the vehicle during the continuous driving of the vehicle after the vehicle switches to the driver-takeover state; taking the center line of the lane in which the vehicle is currently driving as a reference, if the value of the lateral position displacement of the vehicle is positive, it means that the vehicle is deviated laterally to the right during driving; if the value of the lateral position displacement of the vehicle is negative, it means that the vehicle is deviated laterally to the left during driving;

[0023] The experimental process of the automatic driving takeover simulation experiment is as follows: for any automatic driving system takeover scenario, simulate the driving process of the vehicle in the automatic driving state and the road scenario set in the automatic driving system takeover scenario. When the vehicle drives to the set takeover scenario, the automatic driving system of the vehicle fails, and sends a takeover request to the driver according to the interaction method set in the human-computer interaction system. When the driver receives the takeover request, he needs to stop the ongoing driving sub-task and manually control the vehicle to switch from the automatic driving state to the driver takeover state, and obtain the time from the vehicle sending the takeover request to the driver manually controlling the vehicle to switch to the driver takeover state, and use the time as the driver's takeover reaction time; when the takeover is completed, the vehicle automatically switches from the driver takeover state to the automatic driving state, collects the driver's physiological characteristic index data and the vehicle's motion state change data, and ensures that the vehicle's motion state change data always meets the preset experimental conditions; set multiple different driving sub-tasks and repeat the experimental process of the automatic driving takeover simulation experiment to obtain several groups of drivers' takeover reaction time and physiological characteristic index data;

[0024] The step 4 further comprises:

[0025] Step 4.1: Take the takeover reaction time and physiological characteristic index data of each group of drivers as a group of samples;

[0026] Step 4.2: Extract the power spectral density of the Delta band in the driver's EEG signal data;

[0027] Step 4.3: Extracting the heart rate variability (HRV) features from the driver’s ECG signal data;

[0028] Step 4.4: Define physiological weight coefficient;

[0029] Step 4.5: Calculate the driver's physiological weight coefficient according to the driver's Delta band power spectrum density and heart rate variability HRV characteristics;

[0030] Step 4.6: Based on the driver's physiological weight coefficient, the driver's weighted average takeover reaction time DTRT is calculated using the one-way ANOVA method;

[0031] The physiological weight coefficient described in step 4.4 is expressed as:

[0032]

[0033] Where a is the physiological weight coefficient; PSD(f) represents the power spectrum density of the Delta band; f is the frequency; HRV represents the extracted heart rate variability HRV feature; represents the mean of physiological characteristics; represents the standard deviation of physiological characteristics;

[0034] The calculation method of the weighted average takeover reaction time of the driver described in step 4.6 is:

[0035]

[0036] Where DTRT represents the weighted average takeover reaction time of the driver; N is the number of sample groups; x i is the mean of the takeover reaction time in the i-th group of samples, ω i is the weight of the i-th group of samples;

[0037] The calculation method of the vehicle risk collision time in step 5 is:

[0038] RTTC=DTRT+t

[0039] Where RTTC represents the risk collision time of the vehicle; t represents the redundant time.

[0040] The beneficial effects of adopting the above technical solution are:

[0041] The method of the present invention designs a variant decision indicator of TTC, namely, Risk Time to Collision (RTTC). The method of the present invention can more accurately determine when to take over the reminder time through RTTC, so that the driver can have sufficient reaction time to complete the takeover of the vehicle. RTTC analyzes the driver by combining the concepts of takeover reminder and triggering AEB, comprehensively considers many risk factors, and sets the optimal RTTC time to improve the collision risk assessment accuracy of the autonomous driving system in specific situations, thereby providing a more accurate basis for subsequent decision-making, taking into account the complex interaction of man-machine-environment, and proposing a more comprehensive and accurate time window judgment tool for control switching decisions, ensuring the rationality of the decision-making of the autonomous driving system in emergency situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a flow chart of a method for calculating vehicle risk collision time based on expected functional safety in this implementation manner;

[0043] Figure 2 This is a schematic diagram of the risk collision time calculation in this embodiment;

[0044] Figure 3 This is an experimental framework diagram for calculating the risk collision time in this implementation manner. DETAILED DESCRIPTION

[0045] In order to facilitate the understanding of the present application, the specific embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thoroughly understood.

[0046] A method for calculating vehicle risk collision time based on expected functional safety in this embodiment is as follows: Figure 1 and Figure 2 As shown, the method comprises the following steps:

[0047] Step 1: Build an autonomous driving takeover simulation environment, including: scenario simulation module, human-computer interaction system and data acquisition module.

[0048] In this embodiment, in order to conduct a simulated autonomous driving takeover experiment on a driving simulator, it is necessary to first build an autonomous driving system simulation environment. In the L3 autonomous driving takeover process, the three factors of environment, vehicle, and driver are coupled with each other and jointly affect the design of the autonomous driving strategy. Therefore, this embodiment constructs a simulation environment with reference to the above three factors, which is composed of a scene simulation module, a human-computer interaction system, and a data acquisition module.

[0049] The scenario simulation module is used to construct a vehicle dynamics model, takeover control logic and a preset test environment model to simulate the takeover scenario of the automatic driving system.

[0050] The preset test environment model and the human-computer interaction system together constitute a takeover scenario.

[0051] In this embodiment, the scenario simulation module is constructed by joint simulation of VTD (Virtual Test Drive), Simulink and DYNA4, and the constructed simulation scene is displayed on two 27-inch circular LCD screens with a resolution of 1920×1200 pixels. In the VTD-Simulink-dyna4 software joint simulation environment, DYNA4 is used to provide the vehicle's dynamic model, VTD is used to generate scenes and display simulation images, and Simulink is mainly used to manage the switching of autonomous driving and G29 steering wheel takeover logic. The test environment part of the scenario simulation module, that is, the preset test environment model, is mainly constructed by perception data collected by cameras, millimeter-wave radars, lidars and other equipment.

[0052] The human-computer interaction system is used to manage the switching of the vehicle between the automatic driving state and the driver takeover state according to the takeover control logic. When the driver receives a takeover request from the vehicle, the driver manually controls the vehicle to switch from the automatic driving state to the driver takeover state; when the takeover is completed, the vehicle automatically switches from the driver takeover state to the automatic driving state.

[0053] In this embodiment, the vehicle in this experiment is designed to be in automatic driving mode by default, and participants can switch driving control by pressing a button on the steering wheel.

[0054] The human-computer interaction system includes: vehicle status, interaction mode and driver part.

[0055] The vehicle state is used to simulate the real-time driving state of the vehicle during driving; the vehicle state includes: the vehicle speed, vehicle position and steering wheel movement.

[0056] The interaction method is used to simulate the prompting method of the vehicle sending a takeover request to the driver; the interaction method includes: auditory prompting, visual prompting and tactile prompting.

[0057] The driver part is used to simulate different driving states of the driver; the driver part includes: driver attributes and driving sub-tasks; wherein the driver attributes are used to simulate drivers with different driving conditions; the driver attributes include: the driver's age, gender and driving experience; the driving sub-tasks are used to simulate the driver's state when receiving a takeover request; the driving sub-tasks include: talking, sleeping, listening to music and playing games.

[0058] In this embodiment, the human-computer interaction system is mainly composed of the vehicle state, interaction mode and driver part. The vehicle state mainly includes factors such as the vehicle speed, vehicle position and steering wheel. The interaction mode mainly includes three interaction modes: auditory prompts, visual prompts and tactile prompts. The driver part is mainly composed of driver attributes and driving tasks. The driver attributes include the driver's age, gender and driving experience. Driving tasks include talking, sleeping, listening to music and playing games.

[0059] The data acquisition module is used to collect the driver's physiological characteristic index data and the vehicle's motion state change data in a simulated automatic driving system takeover scenario.

[0060] In this embodiment, the data acquisition module is responsible for collecting corresponding data of the driver and the vehicle during the experiment for subsequent analysis.

[0061] The data acquisition module includes: a physiological signal acquisition module and a vehicle motion acquisition module.

[0062] The physiological signal acquisition module is used to collect the driver's electroencephalogram signal data, electrocardiogram signal data, respiratory rate data and skin temperature data in a scenario simulating the automatic driving system taking over.

[0063] In this embodiment, the PhysioLAB multi-channel physiological monitor is used to collect the driver's EEG signal data, ECG signal data, respiratory rate data and skin temperature data during the takeover experiment. Among them, the EEG signal data (Electroencephalogram, EEG) refers to the psychological changes and subtle action decisions of the driver during the driving task, which will be reflected in the brain waves. By monitoring the changes in brain wave indicators, the driver's psychological changes can be objectively and accurately obtained; the electrocardiogram signal data (Electrocardiogram, ECG) is a key indicator for evaluating the driver's physiological state during the automatic driving takeover process, mainly reflecting the activity of the heart; respiratory rate data (Respiratory Rate, RESP): respiratory monitoring is a non-invasive, real-time physiological state assessment method, which can be used to assess the driver's physiological state and readiness to ensure safe takeover; skin temperature (Skin Temperature, TEMP): temperature is an important indicator reflecting the physiological state of the human body. In the human body, due to the excitatory effect of the sympathetic nerves, the sweat gland activity is enhanced, and the skin temperature physiological indicator formed can objectively reflect the activity changes of the sympathetic nervous system.

[0064] The vehicle motion acquisition module is used to collect the longitudinal speed change and lateral offset of the vehicle as the vehicle continues to travel after the vehicle switches to the driver takeover state; wherein the longitudinal speed change of the vehicle is used to record in real time the speed change of the vehicle as the vehicle continues to travel after the vehicle switches to the driver takeover state; the lateral offset of the vehicle is used to record in real time the lateral position offset of the vehicle as the vehicle continues to travel after the vehicle switches to the driver takeover state; taking the center line of the lane in which the vehicle is currently traveling as a reference, if the value of the lateral position offset of the vehicle is positive, it means that the vehicle is laterally offset to the right during driving; if the value of the lateral position offset of the vehicle is negative, it means that the vehicle is laterally offset to the left during driving.

[0065] In this embodiment, the program in the simulation scene is used to record the vehicle speed change and lateral offset during the takeover experiment. The longitudinal speed change of the vehicle is used to indicate the longitudinal control of the vehicle speed by the driver during the takeover process, and the unit is m / s. In the simulated automatic driving system takeover scenario, the driver is required to brake the high-speed vehicle after receiving the takeover request, and the driver's takeover ability is judged by analyzing the vehicle's speed change; the lateral offset of the vehicle refers to the difference between the lateral coordinates of the vehicle and the midpoint of the lane, and the unit is m. Taking the center line of the lane where the vehicle is located during driving as the reference, the data on the left side of the lane center line is set to a negative value, and the data on the right side is set to a positive value, which can be used to reflect the driver's ability to control the lateral movement of the vehicle when taking over.

[0066] Step 2: Set up several groups of different road scenarios, takeover scenarios, and vehicle speeds in the autonomous driving takeover simulation environment, and build several autonomous driving system takeover scenarios.

[0067] The road scenes set in this embodiment are urban expressway scenes and urban road scenes built in the VTD-Simulink-dyna4 software joint simulation environment. The urban expressway scene is 29 kilometers long, with a road speed limit of 100km / h, and the self-driving speed of the vehicle is 80km / h. Two takeover scenarios are designed for this road scene, namely lane line disappearance takeover (C1) and road surface water takeover (C2). The urban road is 32 kilometers long, with a road speed limit of 50km / h, and the self-driving speed of the vehicle is 40km / h. There are two takeover scenarios on this road, namely fog takeover (C3) and rainy day takeover (C4).

[0068] Step 3: Conduct an autonomous driving takeover simulation experiment for each autonomous driving system takeover scenario to obtain takeover reaction time and physiological characteristic indicator data for several groups of drivers.

[0069] In this embodiment, the driver wears a physiological detector and attaches two measuring electrodes of the EEG sensor to specific scalp areas, such as the forehead and occipital bone; attaches the reference electrode to a low muscle activity area, such as the mastoid process behind the ear; attaches the positive electrode of the ECG sensor to the middle position of the line connecting the midpoint of the left clavicle and the chest, and the negative electrode to the fifth intercostal space of the left sternum, and grounds it at the protruding position of the sternum bone. The temperature sensor is attached to the armpit; and the breathing sensor is worn in the abdominal cavity of the human body, that is, the position with greater fluctuations during breathing.

[0070] The experimental process of the automatic driving takeover simulation experiment is as follows: for any automatic driving system takeover scenario, simulate the driving process of the vehicle in the automatic driving state in the road scenario set in the automatic driving system takeover scenario; when the vehicle drives to the set takeover scenario, the automatic driving system of the vehicle fails, and sends a takeover request to the driver according to the interaction method set in the human-computer interaction system; when the driver receives the takeover request, he needs to stop the ongoing driving sub-task and manually control the vehicle to switch from the automatic driving state to the driver takeover state, obtain the time from the vehicle sending the takeover request to the driver manually controlling the vehicle to switch to the driver takeover state, and use the time as the driver's takeover reaction time; when the takeover is completed, the vehicle automatically switches from the driver takeover state to the automatic driving state, collects the driver's physiological characteristic index data and the vehicle's motion state change data, and ensures that the vehicle's motion state change data always meets the preset experimental conditions; set multiple different driving sub-tasks and repeat the experimental process of the automatic driving takeover simulation experiment to obtain several groups of drivers' takeover reaction time and physiological characteristic index data.

[0071] In this embodiment, if Figure 3 As shown, when the vehicle's automatic driving system fails, a takeover request message is sent to the driver in advance according to the set prompt method. After receiving the takeover request message, the driver needs to immediately stop the current driving task and take over urgently by turning the steering wheel or stepping on the brake pedal, that is, the driver manually controls the vehicle to switch from the automatic driving state to the driver takeover state. After the takeover is completed, the automatic driving system is reactivated, and the test is repeated according to the above steps under different driving tasks. The takeover evaluation index is used to evaluate various aspects of the takeover effect and the driver's physiological changes during the takeover process. It is generally divided into driver-related indicators, including: reaction force characteristic indicators, that is, the driver's takeover reaction time, physiological characteristic indicator data and vehicle motion state change data.

[0072] In this embodiment, the takeover reaction time refers to the time it takes for the driver to respond to the takeover request issued by the automatic driving system when the automatic driving system requires the driver to regain control, in seconds. The takeover reaction time is often used to evaluate the speed at which the driver recovers the takeover consciousness and action, and is one of the most widely used indicators for evaluating the driver's takeover reaction characteristics.

[0073] Step 4: Calculate the driver's weighted average takeover reaction time based on the driver's takeover reaction time and physiological characteristic index data.

[0074] Step 4.1: Take the takeover reaction time and physiological characteristic index data of each group of drivers as a set of samples.

[0075] Step 4.2: Extract the power spectral density of the Delta band in the driver's EEG signal data.

[0076] The method for extracting the power spectral density of the Delta band is:

[0077]

[0078] Where PSD(f) represents the power spectral density of the Delta band; f is the frequency; T is the time window length; N is the number of sample groups; n is the nth group of samples; x(n) is the sampling signal of the nth group of samples; e is a natural constant; and j represents an imaginary unit.

[0079] In this embodiment, the Delta band is a component of the EEG brain wave, and its frequency range is between 0.5-4 Hz, which is used to indicate the fatigue level of the driver during the takeover.

[0080] Step 4.3: Extract the heart rate variability (HRV) features from the driver’s ECG signal data.

[0081] The method for extracting the heart rate variability HRV feature is:

[0082]

[0083] Where HRV represents the extracted heart rate variability HRV feature; i represents the i-th group of samples; RR i Indicates the time interval between two consecutive heartbeats; is the average of the heartbeat intervals.

[0084] In this embodiment, HRV features are extracted to reflect the physiological changes of the driver during the takeover period.

[0085] In this embodiment, the higher the power spectrum density, breathing rate and average body temperature of the driver's Delta band, the lower the heart rate variability HRV characteristics, which indicates that the driver is in a state of concentration or tension. Considering the impact of the driver's physiological indicators on the takeover reaction time, a physiological weight coefficient is designed to expand its impact on the takeover reaction time.

[0086] Step 4.4: Define physiological weight coefficients.

[0087] The physiological weight coefficient is expressed as:

[0088]

[0089] Where a is the physiological weight coefficient; represents the mean of physiological characteristics; Represents the standard deviation of the physiological characteristics.

[0090] Step 4.5: Calculate the driver's physiological weight coefficient based on the driver's Delta band power spectrum density and heart rate variability HRV characteristics.

[0091] Step 4.6: Based on the driver's physiological weight coefficient, the driver's weighted average takeover reaction time (DTRT) is calculated using the one-way ANOVA method.

[0092] The calculation method of the weighted average takeover reaction time of the driver is:

[0093]

[0094] Where DTRT represents the weighted average takeover reaction time of the driver; x i is the mean of the takeover reaction time in the i-th group of samples, ω i is the weight of the i-th group of samples; in this embodiment, the inverse of the standard deviation of the takeover reaction time in each group of samples is used as the weight, because the smaller the standard deviation, the more concentrated the data, and the greater the weight.

[0095] In this embodiment, the takeover reaction time means the time from the issuance of the takeover request to the driver successfully taking over and controlling the vehicle to restore stability. A shorter takeover completion time means that the driver can quickly respond and handle the transfer of vehicle control, thereby effectively avoiding traffic accidents in emergency situations. Therefore, as an important part of RTTC, it is crucial to determine the takeover completion time through physiological data and takeover data. In order to study the influence of takeover time and physiological indicators during the takeover experiment, a one-way analysis of variance was used to determine the impact of the automatic driving takeover reaction time on the driver's takeover quality. This embodiment selects the weighted square method, and uses the longest automatic driving takeover reaction time during the experiment as the weight coefficient a to ensure that the obtained DTRT can ensure that all drivers take over the vehicle, which can more reasonably reflect the overall takeover reaction time.

[0096] Step 5: Count the takeover reaction times of all drivers, and select the takeover reaction times that meet the preset ratio from the statistical results as redundant time, and use the weighted average takeover reaction time and redundant time of the drivers to calculate the vehicle risk collision time (Risk Time to Collision, RTTC).

[0097] In this embodiment, RTTC is a key indicator used in the autonomous driving system to determine the driver's takeover request or to activate AEB. The value of RTTC is affected by multiple factors such as man-machine-environment, and is composed of the driver's takeover reaction time (Driver Takeover Reaction Time, DTRT) and the redundant time t, in seconds.

[0098] The calculation method of the vehicle risk collision time is:

[0099] RTTC=DTRT+t (5)

[0100] The redundant time t is a dynamically changing interval. Because the takeover reaction time under different experimental conditions may be affected by multiple factors, all takeover reaction times should be statistically analyzed. According to their distribution, two takeover reaction times that can meet the first 60% and 100% are selected as the corresponding redundant time intervals. This can ensure that most people can complete the takeover behavior within this time period. Finally, the takeover reaction time and the redundant time are combined to obtain the final RTTC.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. A method for calculating vehicle risk collision time based on expected functional safety, characterized in that: The method comprises the following steps: Step 1: Build an autonomous driving takeover simulation environment, including: scenario simulation module, human-computer interaction system and data acquisition module; Step 2: Set up several groups of different road scenarios, takeover scenarios, and vehicle speeds in the autonomous driving takeover simulation environment, and build several autonomous driving system takeover scenarios; Step 3: Conduct an autonomous driving takeover simulation experiment for each autonomous driving system takeover scenario to obtain takeover reaction time and physiological characteristic index data of several groups of drivers; Step 4: Calculate the weighted average takeover reaction time of the driver based on the takeover reaction time and physiological characteristic index data of the driver; Step 5: Count the takeover reaction times of all drivers, and select the takeover reaction times that meet the preset ratio from the statistical results as redundant time, and use the weighted average takeover reaction time and redundant time of the drivers to calculate the vehicle risk collision time.

2. According to claim 1, a method for calculating vehicle risk collision time based on expected functional safety is characterized in that: The scenario simulation module described in step 1 is used to construct a vehicle dynamics model, takeover control logic and a preset test environment model that simulates the takeover scenario of the autonomous driving system; The human-machine interaction system is used to manage the switching of the vehicle between the automatic driving state and the driver takeover state according to the takeover control logic. When the driver receives a takeover request from the vehicle, the driver manually controls the vehicle to switch from the automatic driving state to the driver takeover state; when the takeover is completed, the vehicle automatically switches from the driver takeover state to the automatic driving state; The data acquisition module is used to collect the driver's physiological characteristic index data and the vehicle's motion state change data in a simulated automatic driving system takeover scenario.

3. According to claim 2, a method for calculating vehicle risk collision time based on expected functional safety is characterized in that: The human-machine interaction system includes: vehicle status, interaction mode and driver part; The vehicle state is used to simulate the real-time driving state of the vehicle during driving; the vehicle state includes: vehicle speed, vehicle position and steering wheel movement; The interactive mode is used to simulate the prompting mode of the vehicle sending a takeover request to the driver; the interactive mode includes: auditory prompting, visual prompting and tactile prompting; The driver part is used to simulate different driving states of the driver; the driver part includes: driver attributes and driving sub-tasks; wherein the driver attributes are used to simulate drivers with different driving conditions; the driver attributes include: the driver's age, gender and driving experience; the driving sub-tasks are used to simulate the driver's state when receiving a takeover request; the driving sub-tasks include: talking, sleeping, listening to music and playing games.

4. According to claim 2, a method for calculating vehicle risk collision time based on expected functional safety is characterized in that: The data acquisition module includes: a physiological signal acquisition module and a vehicle motion acquisition module; The physiological signal acquisition module is used to collect the driver's brain electrocardiogram signal data, respiratory rate data and skin temperature data in real time in a scenario where the driver simulates the automatic driving system taking over; The vehicle motion acquisition module is used to collect the longitudinal speed change and lateral offset of the vehicle as the vehicle continues to travel after the vehicle switches to the driver takeover state; wherein the longitudinal speed change of the vehicle is used to record in real time the speed change of the vehicle as the vehicle continues to travel after the vehicle switches to the driver takeover state; the lateral offset of the vehicle is used to record in real time the lateral position offset of the vehicle as the vehicle continues to travel after the vehicle switches to the driver takeover state; taking the center line of the lane in which the vehicle is currently traveling as a reference, if the value of the lateral position offset of the vehicle is positive, it means that the vehicle is laterally offset to the right during driving; if the value of the lateral position offset of the vehicle is negative, it means that the vehicle is laterally offset to the left during driving.

5. The method for calculating vehicle risk collision time based on expected functional safety according to claim 4, characterized in that: The experimental process of the automatic driving takeover simulation experiment is as follows: for any automatic driving system takeover scenario, simulate the driving process of the vehicle in the automatic driving state in the road scenario set in the automatic driving system takeover scenario; when the vehicle drives to the set takeover scenario, the automatic driving system of the vehicle fails, and sends a takeover request to the driver according to the interaction method set in the human-computer interaction system; when the driver receives the takeover request, he needs to stop the ongoing driving sub-task and manually control the vehicle to switch from the automatic driving state to the driver takeover state, obtain the time from the vehicle sending the takeover request to the driver manually controlling the vehicle to switch to the driver takeover state, and use the time as the driver's takeover reaction time; when the takeover is completed, the vehicle automatically switches from the driver takeover state to the automatic driving state, collects the driver's physiological characteristic index data and the vehicle's motion state change data, and ensures that the vehicle's motion state change data always meets the preset experimental conditions; set multiple different driving sub-tasks and repeat the experimental process of the automatic driving takeover simulation experiment to obtain several groups of drivers' takeover reaction time and physiological characteristic index data.

6. The method for calculating vehicle risk collision time based on expected functional safety according to claim 5, characterized in that: The step 4 further comprises: Step 4.1: Take the takeover reaction time and physiological characteristic index data of each group of drivers as a group of samples; Step 4.2: Extract the power spectral density of the Delta band in the driver's EEG signal data; Step 4.3: Extracting the heart rate variability (HRV) features from the driver’s ECG signal data; Step 4.4: Define physiological weight coefficient; Step 4.5: Calculate the driver's physiological weight coefficient according to the driver's Delta band power spectrum density and heart rate variability HRV characteristics; Step 4.6: Based on the driver's physiological weight coefficient, the driver's weighted average takeover reaction time DTRT is calculated using the one-way analysis of variance method.

7. The method for calculating vehicle risk collision time based on expected functional safety according to claim 6, characterized in that: The physiological weight coefficient described in step 4.4 is expressed as: Where a is the physiological weight coefficient; PSD(f) represents the power spectrum density of the Delta band; f is the frequency; HRV represents the extracted heart rate variability HRV feature; represents the mean of physiological characteristics; Represents the standard deviation of the physiological characteristics.

8. The method for calculating vehicle risk collision time based on expected functional safety according to claim 7, characterized in that: The calculation method of the weighted average takeover reaction time of the driver described in step 4.6 is: Where DTRT represents the weighted average takeover reaction time of the driver; N is the number of sample groups; x i is the mean of the takeover reaction time in the i-th group of samples, ω i is the weight of the i-th group of samples.

9. The method for calculating vehicle risk collision time based on expected functional safety according to claim 8, characterized in that: The calculation method of the vehicle risk collision time in step 5 is: RTTC=DTRT+t Where RTTC represents the risk collision time of the vehicle; t represents the redundant time.

Images