A driving right takeover decision method based on human-vehicle-environment safety status perception

Abstract

The present application relates to a driving right takeover decision method based on human-vehicle-environment safety status perception, which belongs to the field of intelligent transportation. The specific steps are: collecting driver parameters to construct a driver's multi-factor takeover capability evaluation model, and obtaining the driver's takeover capability level through the driver parameters; collecting vehicle parameters to construct a vehicle safety boundary and divide the driving safety situation level; constructing a driving right takeover model based on the driver's takeover capability level and the driving safety situation level, and obtaining the takeover result, by integrating the driver's takeover capability and the driving safety situation, and generalizing the parameters of the driver and the vehicle into a description method on the same dimension, so that the two reflect a more accurate and obvious functional relationship compared with the existing technology; constructing a driving right takeover rule model based on the takeover rules, and effectively realizing the driving right takeover strategy allocation in the takeover process under the control goal of ensuring stable and safe driving, and significantly improving the safety and reliability of the automatic driving system.

Description

Technical Field

[0001] The present application belongs to the field of intelligent driving technology, and specifically relates to a driving right takeover decision method based on human-vehicle-environment safety status perception. Background Art

[0002] In recent years, with the continuous development of economy and technology, the number of cars in my country has been rising year by year. The intelligent car technology developed together with the automobile industry is also gradually improving and becoming one of the important development directions in the field of automobile research and development. At present, the mainstream of intelligent car automatic driving level is divided into six levels from L0 to L5 by SAE. Due to the current technical bottlenecks and the complexity of driving tasks during driving, L5 fully automatic driving cars still need a long time to develop. For a long time in the future, automatic driving cars will be in the L3-L4 stage. In this stage, the driving mode of the car will be in the human-machine co-driving mode. In the human-machine co-driving mode, since the automatic driving system has certain applicability and limitations, it involves the takeover of the driving right to the driver under different working conditions. Therefore, it is necessary to identify the driving right takeover conditions. This technology analyzes and evaluates the driving conditions in real time, makes driving right takeover decisions when dangerous conditions occur, and improves driving safety.

[0003] However, there are few studies on the methods of taking over driving rights in the prior art. Among them, the state monitoring and perception of the driver and vehicle during the takeover process are more based on data analysis and prediction, and the range change of the parameters directly corresponds to the state change of the driver or vehicle, which leads to the state step when the parameters change slightly, and the different types of parameters of the two are difficult to describe the problem in the same dimension, lacking generalization methods; secondly, there are few studies on the integration of the driver and the vehicle and the driving environment to take over the driving rights. There is no good mathematical method to couple the three in the method of building a model, lacking comprehensive consideration of the driver-vehicle-driving environment. Summary of the invention

[0004] The purpose of this application is to solve the problem that under the existing technology, in the process of taking over driving rights in a human-machine co-driving environment, it is difficult to integrate multiple factors such as the driver, vehicle, and environment to obtain a comprehensive takeover result, lack of normalization and generalization of parameters in the monitoring of the driver and vehicle, and cannot accurately reflect the actual status of the driver and vehicle and make a takeover decision with a high fault tolerance rate. In response to the defects in the existing technology, this application is based on the safety status of the driver-vehicle-driving environment, a driving right takeover decision method based on the perception of the safety status of the person-vehicle-environment.

[0005] The technical solution of this application is:

[0006] A driving right takeover decision method based on human-vehicle-environment safety state perception includes the following steps:

[0007] Step S1. Collecting driver parameters to construct a driver multi-factor takeover capability assessment model, and obtaining the driver takeover capability level through the driver parameters;

[0008] Step S2. Collect vehicle parameters to construct a vehicle safety boundary, and divide the driving safety situation level based on the vehicle safety boundary;

[0009] Step S3. Based on the driver's takeover capability level described in step S1 and the driving safety situation level described in step S2, a driving right takeover model is constructed to obtain a takeover result.

[0010] Preferably, the driver parameters in step S1 are the driver's eye movement performance and the driver's electroencephalogram signal.

[0011] Preferably, the vehicle parameters in step S2 are the distance the vehicle deviates from the center line of the road and the vehicle kinematic parameters, and the vehicle kinematic parameters are the vehicle speed, the vehicle acceleration and the vehicle heading angle.

[0012] Preferably, the construction of the driver's multi-factor takeover ability assessment model in step S1 is specifically as follows:

[0013] The driver's multi-factor takeover ability assessment model includes two indicators: driver fatigue state FS and driver reaction state RE. The hierarchical analysis method is used to assign indicator weights to the two indicators, and a multi-factor scoring model is constructed to evaluate the driver's takeover ability and divide it into levels.

[0014] Preferably, the driver fatigue state FS is calibrated using the eye closing time ratio in the eye movement performance parameter, as shown in formula (1):

[0015]

[0016] Where Percclos is the percentage of closed eyes time in the eye movement performance parameters, Eye close is the number of eye closure frames per sampling unit time, Eye gamut It is the total number of frames of sampled material within the sampling unit time.

[0017] Preferably, the driver's reaction force state RE is calibrated using the proportion of different bands of the driver's electroencephalogram signal, and the electroencephalogram signal sensor monitors four types of brain waves: α wave, β wave, θ wave, and δ wave.

[0018] Preferably, the specific steps of adopting the hierarchical analysis method are:

[0019] Step 1: Establish a hierarchical structure based on the driver fatigue state FS and the driver reaction state RE, and assign indicator weights of the driver fatigue state FS and the driver reaction state RE to obtain the corresponding judgment matrix of the driver fatigue state FS and the driver reaction state RE;

[0020] Step 2: Perform consistency check on the corresponding judgment matrix of the driver fatigue state FS and the driver reaction state RE in step 1, as shown in the following formulas (2) and (3):

[0021]

[0022]

[0023] Among them, CI is the consistency index, λ is the maximum eigenvalue of the judgment matrix, n is the matrix dimension, CR is the random consistency ratio, and RI is the random consistency index;

[0024] Step 3: After consistency verification, the weight vectors of the driver fatigue state FS and the driver reaction state RE are calculated, and then the weight vectors of the driver fatigue state FS and the driver reaction state RE are multiplied to obtain the final weight distribution;

[0025] Step 4: Based on the driver fatigue state perception results and the driver reaction force perception results, combined with weighting, the driver's takeover ability is calculated through a multi-factor scoring model, as shown in formula (4):

[0026]

[0027] Among them, S i is the comprehensive score of the i-th factor, n is the number of factors, w j is the weight of the jth factor, X ij is the score of the i-th factor on the j-th factor;

[0028] Step 5: Score and grade the driver's takeover ability into five levels: good G, better MG, average M, poor MB, and poor B.

[0029] Preferably, the method of constructing the vehicle safety boundary and dividing the vehicle safety situation level in step S2 is:

[0030] Step S21: Calculate the temporary collision time with the obstacle ahead, as shown in formulas (5) and (6):

[0031] TTCm=TTC+ΔV rel / a (5)

[0032] TTC=Δd / ΔV rel (6)

[0033] Where, TTCm is the time between the improved vehicle and the vehicle in front (obstacle) before collision, TTC is the time before collision, ΔV rel is the relative speed between the vehicle and the obstacle in front, a is the acceleration of the vehicle; Δd ​​is the relative distance between the vehicle and the obstacle in front;

[0034] Step S22: Calculate the distance that the vehicle deviates from the center line, as shown in formula (7):

[0035] LD=P vehicle -P mid (7)

[0036] Among them, LD is the distance that the vehicle deviates from the center line, P vehicle is the lateral position of the vehicle, P mid is the lateral position of the road centerline;

[0037] Step S23. Divide the vehicle's safety boundary based on the temporary collision time calculated in step S21 and the vehicle's deviation distance from the center line calculated in step S22, and further divide the vehicle into driving safety situation levels.

[0038] Preferably, the driving right takeover model in step S3 is constructed based on the following: when the driving safety situation is at a safe or general level, the driving right is handed over to the automatic driving system; when the driving safety situation is dangerous, the driving right is transferred to the driver; when the driving safety situation is dangerous and the driver's takeover ability level is low, the system makes the driving decision.

[0039] Preferably, the method for constructing the driving right takeover model in step S3 is: driver The safety level is set to five levels: good G, better MG, average M, poor MB, and poor B; vehicle It is set into three levels: safe S, general M, and dangerous B, and outputs the takeover results in different safety status scenarios.

[0040] Beneficial effects of this application:

[0041] This application integrates the driver's takeover capability and driving safety situation, and generalizes the parameters of the driver and the vehicle into a description method on the same dimension, so that the two reflect a more accurate and obvious relationship of action compared with the existing technology; in addition, considering the vehicle driving safety environment, the driving right takeover rule strategy is designed, and the driving right takeover rule model is constructed based on the takeover rule, which can realize the relationship coupling between people, vehicles and the environment under an accurate mathematical model, and can obtain the takeover result more accurately than the existing technology, and effectively realize the allocation of driving right takeover strategy in the takeover process under the control goal of ensuring stable and safe driving. It can significantly improve the safety and reliability of the automatic driving system and reduce the risk of traffic accidents.

[0042] This application can evaluate the driver's fatigue and reaction state in real time by collecting driver parameters, such as eye movement performance and EEG signals, so as to build a multi-factor driver takeover ability evaluation model. This model can comprehensively consider the driver's safety status, realize intelligent driving right takeover decision-making, and improve the safety and reliability of the autonomous driving system.

[0043] This application can construct the vehicle safety boundary and classify the vehicle safety situation level by collecting vehicle parameters, such as the distance the vehicle deviates from the road centerline and the vehicle kinematic parameters. It can evaluate the safety status of the vehicle in real time, providing another important basis for the decision to take over the driving rights.

[0044] This application uses the hierarchical analysis method to assign weights to driver parameters, which can more scientifically and reasonably evaluate the driver's takeover ability; the driver's takeover ability is divided into five takeover ability levels: excellent G, good MG, general M, poor MB, and poor B, which can more clearly reflect the driver's status during vehicle driving and improve driving safety.

[0045] The driving right takeover decision method based on human-vehicle-environment safety status perception of the present application is highly intelligent and adaptive, and can make reasonable driving right takeover decisions according to the safety status of the driver and the vehicle, providing strong support for the development and application of autonomous driving technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a technical roadmap for the driving rights takeover method for this application.

[0047] Figure 2 A flowchart of the driver's multi-factor takeover ability assessment model was constructed for this application.

[0048] Figure 3 A model flow chart of driving rights takeover was constructed for this application. DETAILED DESCRIPTION

[0049] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific implementation method, structure, characteristics and effects of the present application are described in detail below in combination with the accompanying drawings and preferred embodiments.

[0050] Example 1

[0051] The present embodiment provides a data collection device, which includes a driver perception subsystem and a vehicle perception subsystem. Specifically, the driver perception subsystem includes a driver parameter collector, a driver parameter communicator, and a driver parameter processor. The driver parameter collector includes a wearable eye tracker and a driver electroencephalogram signal sensor. The driver parameter communicator adopts a CAN communication method. The driver parameter processor is a software / hardware device supporting the parameter collector. The vehicle perception subsystem includes a vehicle position sensor, a vehicle camera, a vehicle speed sensor, a vehicle acceleration sensor, and a vehicle data processor. The composition of the perception subsystem and the collection parameters are shown in Table 1.

[0052] Table 1 Composition and acquisition parameters of the perception subsystem

[0053]

[0054]

[0055] Example 2

[0056] This embodiment provides a driving right takeover decision method based on human-vehicle-environment safety state perception, and the specific steps are as follows: Figure 1 As shown, the following steps are included:

[0057] Step S1. Collecting driver parameters to construct a driver multi-factor takeover capability assessment model, and obtaining the driver takeover capability level through the driver parameters;

[0058] Among them, the driver perception subsystem is used to collect driver parameters, which include the driver's eye movement performance and EEG signals.

[0059] Construct a multi-factor takeover ability assessment model for drivers, such as Figure 2 As shown, specifically:

[0060] The driver's multi-factor takeover ability assessment model includes two indicators: driver fatigue state FS and driver reaction state RE. The hierarchical analysis method is used to assign indicator weights to the two indicators, and a multi-factor scoring model is constructed to evaluate the driver's takeover ability and divide it into levels.

[0061] The driver's fatigue state FS is calibrated using the eye-closing time ratio in the eye movement performance parameter, as shown in the following formula (1):

[0062]

[0063] In the formula, Percclos is the proportion of the closed-eye time in the eye movement performance parameters, Eye close is the number of frames with the eyes closed within the sampling unit time of the eye tracker, Eye gamut is the total number of frames of the sampling material within the sampling unit time of the eye tracker. The fatigue state FS of the driver is determined by the value of Percclos. When 0 < Percclos ≤ 0.04, the driver is determined to be in a sober state. When 0.04 ≤ Percclos < 0.06, the driver is determined to be in a relatively sober state. When 0.06 ≤ Percclos < 0.09, the driver is determined to be in a general fatigue state. When 0.09 ≤ Percclos < 0.12, the driver is in a relatively fatigued state. When Percclos ≥ 0.12, the driver is determined to be in a fatigued state. The classification of the driver's fatigue state is shown in Table 2 below:

[0064] Table 2 Classification of Driver Fatigue State

[0065]

[0066] The driver's reaction force state RE is calibrated by the proportion of different frequency bands of the driver's electroencephalogram (EEG) signals. The EEG signal sensor monitors four types of brain waves: alpha wave (8 - 13 hz, with a higher proportion of alpha waves appearing when a person is in a relaxed state), beta wave (above 13 hz, with a higher proportion of beta waves appearing when a person is in a focused state), theta wave (4 - 7 hz, with a higher proportion of theta waves appearing when a person is in a slack state), and delta wave (1 - 4 hz, with a higher proportion of delta waves appearing when a person is in a sleeping state). It is set that the sum of the four waveforms is 100%. On this basis, when the proportion of alpha waves and beta waves appears greater than 50%, and the proportion of theta waves and delta waves appears less than 30%, it is determined that the driver is in a high reaction force state. When the proportion of alpha waves and beta waves appears greater than 50%, and the proportion of theta waves and delta waves appears greater than or equal to 30% and less than or equal to 50%, it is determined that the driver is in a relatively high reaction force state. When the proportion of alpha waves and beta waves appears greater than or equal to 30% and less than or equal to 50%, and the proportion of theta waves and delta waves appears greater than or equal to 30% and less than or equal to 50%, it is determined that the driver is in a general reaction force state. When the proportion of alpha waves and beta waves appears greater than or equal to 30% and less than or equal to 50%, and the proportion of theta waves and delta waves appears greater than 50%, it is determined that the driver is in a relatively low reaction force state. When the proportion of alpha waves and beta waves appears less than 30%, and the proportion of theta waves and delta waves appears greater than 50%, it is determined that the driver is in a low reaction force state. The classification of the driver's reaction force state is shown in Table 3 below.

[0067] Table 3

[0068]

[0069] By collecting driver parameters such as eye movement performance and EEG signals, the driver's fatigue state FS and reaction state RE can be evaluated in real time, thereby building a driver's multi-factor takeover ability evaluation model. This model can comprehensively consider the driver's safety status, realize intelligent driving right takeover decision-making, and improve the safety and reliability of the autonomous driving system.

[0070] The specific steps of using the hierarchical analysis method are as follows:

[0071] Step 1: Establish a hierarchical structure based on the driver's fatigue state FS and the driver's reaction state RE, and assign indicator weights to the driver's fatigue state FS and the driver's reaction state RE. According to expert experience, the following criteria are set: the sum of the weights of the driver's fatigue state FS and the reaction state RE is 1. Compared with the reaction state, the fatigue state has a more significant impact on the takeover ability. Therefore, the fatigue state FS is assigned an indicator weight of 0.6, and the reaction state RE is assigned an indicator weight of 0.4, and the corresponding judgment matrix of the driver's fatigue state FS and the driver's reaction state RE is obtained;

[0072] Step 2: Perform consistency check on the corresponding judgment matrix of the driver fatigue state FS and the driver reaction state RE in step 1, as shown in the following equations (2) and (3):

[0073]

[0074]

[0075] Among them, CI is the consistency index, λ is the maximum eigenvalue of the judgment matrix, n is the matrix dimension, CR is the random consistency ratio, and RI is the random consistency index;

[0076] Step 3: After consistency verification, the weight vectors of the driver fatigue state FS and the driver reaction state RE are calculated, and then the weight vectors of the driver fatigue state FS and the driver reaction state RE are multiplied to obtain the final weight distribution;

[0077] Step 4: Based on the driver fatigue state perception results and the driver reaction force perception results, combined with weighting, the driver's takeover ability is calculated through a multi-factor scoring model, as shown in the following formula (4):

[0078]

[0079] Among them, S i is the comprehensive score of the i-th factor, n is the number of factors, w j is the weight of the jth factor, X ij is the score of the i-th factor on the j-th factor;

[0080] Step 5: Score and grade the driver's takeover ability into five levels: excellent G, good MG, average M, poor MB, and poor B.

[0081] By using the hierarchical analysis method to assign weights to driver parameters, the driver's takeover ability can be evaluated more scientifically and reasonably; dividing the driver's takeover ability into five levels: excellent G, good MG, general M, poor MB, and poor B, can more clearly reflect the driver's status during vehicle driving and improve driving safety.

[0082] Step S2. Collect vehicle parameters to construct vehicle safety boundaries and classify vehicle safety situation levels;

[0083] Among them, the vehicle parameters collected by the vehicle perception subsystem are the distance the vehicle deviates from the road centerline and the vehicle kinematic parameters. The vehicle kinematic parameters are the vehicle speed, the vehicle acceleration and the vehicle heading angle.

[0084] The method for constructing the vehicle safety boundary and dividing the vehicle safety situation level in step S2 is:

[0085] Step S21: Calculate the temporary collision time with the obstacle ahead, as shown in the following equations (5) and (6):

[0086] TTCm=TTC+ΔV rel / a (5)

[0087] TTC=Δd / ΔV rel (6)

[0088] Where, TTCm is the time between the improved vehicle and the vehicle in front (obstacle) before collision, TTC is the time before collision, ΔV rel is the relative speed between the vehicle and the obstacle in front, a is the acceleration of the vehicle; Δd ​​is the relative distance between the vehicle and the obstacle in front;

[0089] Step S22: Calculate the distance that the vehicle deviates from the center line, as shown in formula (7):

[0090] LD=P vehicle -P mid (7)

[0091] Among them, LD is the distance that the vehicle deviates from the center line, P vehicle is the current vehicle lateral position, P mid is the lateral position of the road centerline,

[0092] Step S23. Different safety situation levels are divided based on the temporary collision time calculated in step S21 and the distance of the vehicle from the center line calculated in step S22. When the distance of the vehicle from the center line of the road is less than or equal to 1.2m and the temporary collision time is greater than or equal to 5s, the vehicle is determined to be on the safety boundary or within the safety boundary; when any parameter does not meet the above conditions, the vehicle is determined to be not within the safety boundary and the safety level is reduced, and the numerical ranges of the above two parameters are set to be within the safety boundary, on the safety boundary, and outside the safety boundary, respectively, and divided into three driving safety situation levels: safe (S), general (M), and dangerous (B).

[0093] By collecting vehicle parameters, such as the distance the vehicle deviates from the road centerline and vehicle kinematic parameters, the vehicle safety boundary can be constructed and the vehicle safety situation level can be divided. The vehicle safety status can be evaluated in real time, providing another important basis for the decision to take over the driving rights.

[0094] Step S3. Construct a driving right takeover model based on the driver's takeover capability level in step S1 and the vehicle safety situation level in step S2. Figure 3 As shown, the takeover result is obtained.

[0095] The construction basis of the driving right takeover model is as follows: when the driving safety situation is safe or general, it means that the automatic driving system is competent for the driving operation at this time. In order to reflect the role of the automatic driving system and reduce the driver's operating load, the driving right is handed over to the automatic driving system, that is, the driving right belongs to the vehicle, which is represented as "0" in the takeover rule result (E); on the contrary, if the driving safety situation is dangerous at this time, in the human-machine co-driving environment, the automatic driving system is difficult to be competent for the driving scene at this time, then the driving right should be transferred to the driver in time, that is, a takeover request occurs, and the driving right belongs to the driver, which is represented as "1" in the takeover rule result (E); in addition, if the driving safety situation is dangerous, but the driver's takeover ability level is also very low at this time, it means that the driver is in a state that is difficult to complete the control of the vehicle in a dangerous state, so in this case, the driving right is still kept in the automatic driving system, and the system makes driving decisions.

[0096] The construction method of the driving right takeover model is as follows: first, the driver's takeover ability A driver The safety situation is divided into five levels: excellent (G), good (MG), fair (M), poor (MB), and bad (B); secondly, according to the threshold setting of the safety boundary, the vehicle safety situation level A is divided into inside the boundary, above the boundary, and outside the boundary. vehicleIt is set to three levels: safe (S), general (M), and dangerous (B); the output variable takeover result (E) is divided into two results: the driving right belongs to the driver "1", and the driving right belongs to the vehicle control system "0". It is verified in two scenarios with different safety levels, and combined with the input and output variables to obtain: As shown in Tables 4 and 5 below, in different safety status scenarios, the driver's takeover ability level A driver and driving safety level A vehicle 3*5 takeover strategy result table.

[0097] Table 4 Takeover results under straight-ahead emergency obstacle avoidance conditions

[0098]

[0099] Table 5. Take-over results under normal driving conditions at a constant speed

[0100]

[0101]

[0102] In summary, this application integrates the driver's takeover capability and driving safety situation, and generalizes the parameters of the driver and the vehicle into a description method on the same dimension, so that the two reflect a more accurate and obvious relationship of action compared with the existing technology; in addition, considering the vehicle driving safety environment, the driving right takeover rule strategy is designed, and the driving right takeover rule model is constructed based on the takeover rule, which can realize the relationship coupling between people, vehicles and the environment under an accurate mathematical model, and can obtain the takeover result more accurately than the existing technology, and effectively realize the allocation of driving right takeover strategy in the takeover process under the control goal of ensuring stable and safe driving. It can significantly improve the safety and reliability of the automatic driving system and reduce the risk of traffic accidents.

[0103] The above is only a preferred embodiment of the present application and does not constitute any form of limitation to the present application. Although the present application has been disclosed as a preferred embodiment as above, it is not intended to limit the present application. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present application. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present application without departing from the content of the technical solution of the present application are still within the scope of the technical solution of the present application.

Claims

1. A driving right takeover decision method based on human-vehicle-environment safety status perception, characterized in that: The steps include: Step S1. Collecting driver parameters to construct a driver multi-factor takeover capability assessment model, and obtaining the driver takeover capability level through the driver parameters; The construction of the driver's multi-factor takeover ability assessment model in step S1 is specifically as follows: The multi-factor takeover ability assessment model for drivers includes two indicators: driver fatigue state FS and driver reaction state RE. The hierarchical analysis method is used to assign indicator weights to the two indicators, and a multi-factor scoring model is constructed to assess the driver's takeover ability and classify it. The specific steps of using the hierarchical analysis method are: Step 1: Establish a hierarchical structure based on the driver fatigue state FS and the driver reaction state RE, and assign indicator weights of the driver fatigue state FS and the driver reaction state RE to obtain the corresponding judgment matrix of the driver fatigue state FS and the driver reaction state RE; Step 2: Perform consistency check on the corresponding judgment matrix of the driver fatigue state FS and the driver reaction state RE in step 1, as shown in the following formula: Among them, CI is the consistency index, λ is the maximum eigenvalue of the judgment matrix, n is the matrix dimension, CR is the random consistency ratio, and RI is the random consistency index; Step 3: After consistency verification, the weight vectors of the driver fatigue state FS and the driver reaction state RE are calculated, and then the weight vectors of the driver fatigue state FS and the driver reaction state RE are multiplied to obtain the final weight distribution; Step 4: Based on the driver fatigue state perception results and the driver reaction force perception results, combined with weighting, the driver's takeover ability is calculated through a multi-factor scoring model, as shown in the following formula: Among them, S i is the comprehensive score of the i-th factor, n is the number of factors, w j is the weight of the jth factor, X ij is the score of the i-th factor on the j-th factor; Step 5: Score and grade the driver's takeover ability into five levels: good G, better MG, average M, poor MB, and poor B; Step S2. Collect vehicle parameters to construct a vehicle safety boundary, and divide the driving safety situation level based on the vehicle safety boundary; Step S3. Based on the driver's takeover capability level described in step S1 and the driving safety situation level described in step S2, a driving right takeover model is constructed to obtain a takeover result.

2. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The driver parameters in step S1 are the driver's eye movement performance and the driver's electroencephalogram signal.

3. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The vehicle parameters in step S2 are the distance that the vehicle deviates from the center line of the road and the vehicle kinematic parameters, and the vehicle kinematic parameters are the vehicle speed, the vehicle acceleration and the vehicle heading angle.

4. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The driver fatigue state FS is calibrated by using the eye closing time ratio in the eye movement performance parameter, as shown in the following formula: Where Percclos is the percentage of closed eyes time in the eye movement performance parameters, Eye close is the number of eye closure frames per sampling unit time, Eye gamut It is the total number of frames of sampled material within the sampling unit time.

5. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The driver's reaction state RE is calibrated by using the proportion of different bands of the driver's electroencephalogram signal. The electroencephalogram signal sensor monitors four types of brain waves: α wave, β wave, θ wave, and δ wave.

6. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The method for constructing the vehicle safety boundary and dividing the vehicle safety situation level in step S2 is: Step S21. Calculate the temporary collision time with the front obstacle, as shown in the following formula: TTCm=TTC+ΔV rel / a TTC=Δd / ΔV rel Where, TTCm is the time between the improved vehicle and the vehicle in front (obstacle) before collision, TTC is the time before collision, ΔV rel is the relative speed between the vehicle and the obstacle in front, a is the acceleration of the vehicle; Δd ​​is the relative distance between the vehicle and the obstacle in front; Step S22: Calculate the distance that the vehicle deviates from the center line, as shown in the following formula: LD=P vehicle -P mid Among them, LD is the distance that the vehicle deviates from the center line, P vehicle is the lateral position of the vehicle, P mid is the lateral position of the road centerline; Step S23. Divide the vehicle's safety boundary based on the temporary collision time calculated in step S21 and the vehicle's deviation distance from the center line calculated in step S22, and further divide the vehicle into driving safety situation levels.

7. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The driving right takeover model in step S3 is constructed based on the following: when the driving safety situation is at a safe or general level, the driving right is handed over to the automatic driving system; when the driving safety situation is dangerous, the driving right is transferred to the driver; when the driving safety situation is dangerous and the driver's takeover ability level is low, the system makes the driving decision.

8. The driving right takeover decision method based on human-vehicle-environment safety status perception according to claim 1 is characterized in that: The method for constructing the driving right takeover model in step S3 is: driver The safety level is set to five levels: good G, better MG, average M, poor MB, and poor B; vehicle It is set into three levels: safe S, general M, and dangerous B, and outputs the takeover results in different safety status scenarios.

Images