Fatigue monitoring and warning system, method, device and storage medium

Abstract

The present application belongs to the technical field of fatigue driving detection, and particularly relates to a fatigue monitoring and early warning system, method, device and storage medium, comprising a domain controller and a man-machine interaction system, an intelligent front camera, an electric power steering system and a capacitive steering wheel in communication connection with the domain controller; the domain controller is used for receiving state information detected by the intelligent front camera, the electric power steering system and the capacitive steering wheel, and judging the fatigue degree of the driver according to the state information; the man-machine interaction system is used for receiving the instruction of the domain controller and sending a fatigue alarm signal. The present application can realize fatigue monitoring and early warning of the driver without collecting the facial information of the driver, and meets the requirements of relevant regulations in various countries and regions.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of fatigue driving detection, and particularly relates to a fatigue monitoring and early warning system, method, device and storage medium. BACKGROUND

[0002] With the development of intelligent driving technology and vehicle networking, the requirements for vehicle driving safety are becoming higher and higher. Some countries and regions have introduced a series of relevant laws, and it is mandatory to install a system with certain driving safety related functions in vehicles. For example, the general safety regulation (GSR regulation) of the European Union stipulates that vehicles must have a DDAW (driver drowsiness and attention warning) system, which requires the vehicle to monitor the drowsiness level of the driver and issue a warning when the driver is driving while tired.

[0003] For example, the disclosure number CN112863128A discloses a fatigue driving recognition system based on facial features, relating to the field of fatigue driving prevention, which comprises a field of view collection device for collecting field of view image data of the driver's facing position; a face collection device for collecting face image data of the driver; a control device arranged in the driver's cabin and connected with the field of view collection device and the face collection device, for controlling the field of view collection device to collect the field of view image data when the vehicle is driving; when the field of view image data does not change within a preset time range, the control device generates a first collection instruction to control the face collection device to collect the face image data and input a preset model; when the face image data does not match the preset face template, the preset model outputs a recognition result of fatigue driving. The present application effectively improves the accuracy of fatigue driving recognition through double judgment of field of view image data and face image data.

[0004] For example, the disclosure number CN108001452A discloses a fatigue driving prevention system based on facial recognition technology, which comprises a facial recognition system installed on the rearview mirror inside the car, the camera of the facial recognition system matches the driver's seat, the facial recognition system is connected with the processor in the car, and the processor is also installed with a database, a gyroscope module and a tactile interaction module; the present system is based on facial recognition technology, and real-time collection of driver's face parameters and vehicle motion parameters is carried out during driving, the processor judges the driving state of the driver according to the vehicle motion parameters, matches the driving state of the driver with the face parameters, and when the driving state matched by the face parameters of the driver is fatigue driving, the processor sends a trigger signal to the tactile interaction module to control the action of the tactile interaction module, regardless of whether the motion parameters of the vehicle belong to fatigue driving. The problem that the existing auxiliary system generally prompts the driver through the instrument panel or digital display screen and is easily ignored by the driver is solved.

[0005] In the prior art, most of the existing driver fatigue monitoring products on the market use in-cabin cameras to collect the facial information of the driver for monitoring, and judge the fatigue degree of the driver by monitoring the blink frequency, mouth closing degree and other facial features of the driver. However, with the gradual standardization of privacy and data protection requirements, the in-cabin camera products, the form of personal data collected by the in-cabin camera products, and the storage and use requirements of the related data are becoming increasingly stringent. SUMMARY

[0006] In view of the above problems, the present application aims to provide a fatigue monitoring and warning system, method, device and storage medium.

[0007] The present application provides the following technical solutions:

[0008] A fatigue monitoring and warning system, comprising a domain controller and a human-computer interaction system, an intelligent front camera, an electric power steering system and a capacitive steering wheel connected in communication with the domain controller.

[0009] The domain controller is configured to receive state information detected by the intelligent front camera, the electric power steering system and the capacitive steering wheel, and determine the fatigue degree of the driver based on the state information.

[0010] The human-computer interaction system is configured to receive instructions from the domain controller and send a fatigue warning signal.

[0011] The intelligent front camera is configured to detect vehicle driving information, which includes road information, unintended line pressing and lane center deviation driving behavior.

[0012] The electric power steering system is configured to detect state information of the steering wheel, which includes the hand torque applied by the driver on the steering wheel, the steering wheel rotation angle and the steering wheel rotation angular velocity.

[0013] The capacitive steering wheel is configured to detect the force on the steering wheel and determine whether the driver has released the steering wheel. The force on the steering wheel is the grip force applied by the driver on the steering wheel.

[0014] The road information includes lane lines and road signs.

[0015] Let the initial fatigue degree level be f, then f=W i1 X1+W i2 X2+W i3 X3+θ;

[0016] W i1 , W i2 , and W i3 are the weight coefficients of X1, X2 and X3, respectively.

[0017] θ is a bias, which is obtained by fitting the test results of the real vehicle with the theoretical results;

[0018] X1 is the number of times of the vehicle's unintended line pressure and lane center driving behavior within the period detected by the intelligent front camera;

[0019] X2=K1P+K2Q;

[0020] K1 and K2 are weight coefficients of P and Q, respectively;

[0021] P is the proportion of time that the hand torque on the steering wheel exceeds the preset range of hand torque within the period detected by the electric power steering system;

[0022] Q is the proportion of time that the steering wheel rotation angle and the steering wheel rotation angular velocity exceed the compliance interval within the period; the compliance interval of the steering wheel rotation angle and the steering wheel rotation angular velocity is obtained based on the fitting curve of the steering wheel rotation angle and the steering wheel rotation angular velocity during normal driving and by setting a tolerance;

[0023] X3 is the number of times of hand-off within the period;

[0024] f is used to match the fatigue level table to determine the corresponding fatigue level.

[0025] The fatigue level table is a KSS drowsiness table.

[0026] The domain controller is also used to obtain the data of the driver during normal driving as a training set, and adjust the judgment of the driver's fatigue degree based on a neural network algorithm.

[0027] The driver's fatigue degree level is defined as F, F=R1f+R2f’+Ω;

[0028] f’ is the fatigue degree level calculated by learning and training based on the data during normal driving, and its calculation method is the same as f;

[0029] Ω is a set bias;

[0030] R1 and R2 are weight coefficients of f and f’, respectively, R1 and R2 are obtained by a feedforward neural network algorithm, and R 1+ R2≤1, R2 R1.

[0031] F is used to match the range corresponding to each level of the fatigue level table, and trigger the fatigue warning signal of the corresponding level.

[0032] A fatigue monitoring and warning method, comprising the following steps:

[0033] S1, define the range of hand torque applied by the driver to the steering wheel in normal driving conditions; fit the steering wheel rotation angle and steering wheel rotation angle velocity curve in normal driving, and set the tolerance as the compliance interval of the steering wheel rotation angle and steering wheel rotation angle velocity in normal driving;

[0034] Determine the range corresponding to each level of the fatigue level table, and set the corresponding fatigue warning signal;

[0035] S2, the intelligent front camera detects road information, unexpected line pressing and off-center lane driving behavior, and outputs the number of times X1 of unexpected line pressing and off-center lane driving behavior of the vehicle in a specified period to the domain controller in real time;

[0036] The electric power steering system detects the hand torque applied by the driver to the steering wheel, the steering wheel rotation angle, and the steering wheel rotation angle velocity, and outputs the time proportion P of the hand torque applied by the driver to the steering wheel exceeding the set normal driving hand torque range, and the time proportion Q of the steering wheel rotation angle and the steering wheel rotation angle velocity exceeding the compliance interval in a specified period to the domain controller in real time;

[0037] The capacitive steering wheel detects the grip force applied by the driver to the steering wheel, judges whether to release the hand, and outputs the number of times X3 of the driver releasing the hand in a specified period to the domain controller in real time;

[0038] S3, define the initial driver fatigue level as f, then f=W i1 X1+W i2 X2+W i3 X3+θ;X2=K1P+K2Q;

[0039] Match f with the range corresponding to each level of the fatigue level table, and send the corresponding fatigue warning signal instruction to the human-computer interaction system;

[0040] S4, after receiving the instruction of the domain controller, the human-computer interaction system issues a fatigue warning signal.

[0041] In S3, the domain controller also adjusts the judgment of the driver's fatigue degree by taking the driver's operation habit as a training set based on the neural network algorithm;

[0042] Define the driver fatigue level as F, F=R1f+R2f’+Ω;

[0043] f' is the fatigue level calculated according to the data in normal driving, and its calculation method is the same as f;

[0044] Match F with the range corresponding to each level of the fatigue level table, and send the corresponding fatigue warning signal instruction to the human-computer interaction system.

[0045] A fatigue monitoring and early warning device, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the fatigue monitoring and early warning method described above when executing the program.

[0046] A computer-readable storage medium having a computer program stored thereon, wherein the program is executed by a processor to implement the fatigue monitoring and early warning method described above.

[0047] The present application has the following advantages: the present application provides a domain controller, a man-machine interaction system in communication connection with the domain controller, a smart front camera, an electric power steering system and a capacitive steering wheel, detects various lane lines, road signs and other road information through the smart front camera, detects whether there is a lane line pressing behavior or a lane center line deviating behavior against the driver's will according to the actual driving track of the vehicle and the actual operation of the driver, detects the hand torque, the steering wheel turning angle, the steering wheel turning speed and other information of the driver exerted on the steering wheel through the electric power steering system, detects the grip of the driver exerted on the steering wheel through the capacitive steering wheel to determine whether the driver has let go of the steering wheel, and then the domain controller receives the information detected by the smart front camera, the electric power steering and the capacitive steering wheel to determine the driving behavior of the driver, compares it with the standard determined in advance, records the driving habits of the driver for continuous learning, and comprehensively judges the fatigue degree of the driver at the moment. Therefore, the driver's facial features do not need to be collected to determine the fatigue degree of the driver, and the privacy is protected. Meanwhile, the present application can learn deeply from the driving habits of the driver, take the operation habits of the driver as a training set, adjust the corresponding relationship between the calibrated driving behavior and fatigue level by a certain proportion based on a neural network algorithm, so that the performance of the system is more in line with the operation habits of the driver and more humanized. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 It is a system schematic diagram of the present application.

[0049] In the figure, the marks are: smart front camera 101, electric power steering system 102, capacitive steering wheel 103, domain controller 104, man-machine interaction system 105. DETAILED DESCRIPTION

[0050] Embodiment one

[0051] As Figure 1 shown, a fatigue monitoring and early warning system, comprising a domain controller 104 and a man-machine interaction system 105 in communication connection with the domain controller 104, a smart front camera 101, an electric power steering system 102 and a capacitive steering wheel 103, i.e. each module is connected through a connector and a wire harness, and the communication between the modules is realized by using a vehicle-mounted bus.

[0052] The domain controller 104 is configured to receive the state information detected by the intelligent front camera 101, the electric power steering system 102, and the capacitive steering wheel 103, and determine the fatigue degree of the driver based on the state information. The human-computer interaction system 105 is configured to receive the instructions from the domain controller and send a fatigue warning signal.

[0053] The intelligent front camera 101 is configured to detect the vehicle driving information. The vehicle driving information includes road information, unintended lane pressing, and off-center lane driving behavior.

[0054] Specifically, the intelligent front camera 101 is installed on the inner side of the windshield of the vehicle and located above the center axis, and the camera faces the front of the vehicle to detect the road information such as lane lines, road signs, etc., and detect whether there is an unintended lane pressing, off-center lane driving, etc. behavior according to the actual driving trajectory of the vehicle and the actual operation of the driver.

[0055] The electric power steering system 102 is configured to detect the state information of the steering wheel, and the state information of the steering wheel includes the hand torque applied by the driver on the steering wheel, the steering wheel rotation angle, and the steering wheel rotation angular velocity.

[0056] The capacitive steering wheel 103 is configured to detect the force on the steering wheel and determine whether the hand is off. The force on the steering wheel is the grip force applied by the driver on the steering wheel.

[0057] The intelligent front camera 101 detects in real time whether there is an unintended lane pressing and off-center lane driving behavior during the driving of the vehicle, defines the number of unintended lane pressing and off-center lane driving behaviors of the vehicle within a specified period as X1, and outputs X1 to the domain controller 104 in real time.

[0058] The electric power steering system 102 detects in real time the hand torque applied by the driver on the steering wheel, the steering wheel rotation angle, and the steering wheel rotation angular velocity. P is the time proportion of the hand torque on the steering wheel exceeding the preset range of the hand torque within the detection period of the electric power steering system. Specifically, the range of the hand torque applied by the driver on the steering wheel under normal driving conditions is defined, and the time proportion of the hand torque exceeding the set range within a specified period is counted as P. Q is the time proportion of the steering wheel rotation angle and the steering wheel rotation angular velocity exceeding the compliance interval within the period. Specifically, the compliance interval of the steering wheel rotation angle and the steering wheel rotation angular velocity is obtained based on the fitting curve of the steering wheel rotation angle and the steering wheel rotation angular velocity under normal driving conditions and a set tolerance. The electric power steering system outputs X2=K1P+K2Q. K1 and K2 are weight coefficients of P and Q, respectively, which can be set according to the specific algorithm. The electric power steering system 102 sends X2 to the domain controller 104 in real time.

[0059] The capacitive steering wheel 103 detects the grip force applied by the driver on the steering wheel in real time, and judges whether the driver is off-hand. Define the number of off-hand times in a specified period as X3. The capacitive steering wheel 103 outputs X3 to the domain controller 104 in real time.

[0060] The domain controller 104 receives the state information sent by the intelligent front camera 101, the electric power steering system 102, and the capacitive steering wheel 103 in real time, and comprehensively calculates the fatigue degree of the driver at this moment through the information such as the frequency of pressing the line, the frequency of driving off the center of the lane, the change of the grip force applied by the driver on the steering wheel, whether the driver is off-hand driving, the size of the hand torque applied on the steering wheel when steering, the size of the steering angle when steering, and whether it is sudden steering, etc. recorded in a certain time period.

[0061] Specifically, define the initial driver fatigue level as f, then f = W i1 X1 + W i2 X2 + W i3 X3 + θ.

[0062] θ is a bias, which is obtained by fitting the results of real vehicle test and theoretical results.

[0063] W i1 , W i2 , and W i3 are the weight coefficients of X1, X2, and X3 respectively, which can be obtained through actual experience calibration during the development process, or through real vehicle driving training by BP neural network algorithm.

[0064] W i1 , W i2 , and W i3 need to be obtained according to a certain amount of real vehicle test. The people participating in the test need to complete the relevant training of KSS, and can judge the corresponding drowsiness level according to their own state (some medical instruments can also be used to assist in judging the drowsiness level at that time), while recording the driving performance under the corresponding drowsiness level. According to the actual performance statistics under each level, the influence of X1, X2, and X3 can be determined. For example, assuming that when the drowsiness level is 8, the range of X1 collected is 10~15, and the range of X3 is 2~3, then W i1 should be adjusted to be smaller in proportion, so as to avoid the influence of X1 times on f being too large.

[0065] f is used to match the range corresponding to each level of the fatigue level table to determine the corresponding fatigue level. The fatigue level table can be selected as the KSS drowsiness table.

[0066] For example, the GSR regulation stipulates that a warning can be issued at level 7; a warning must be issued at level 8 / 9. We can set that when the drowsiness level reaches level 7, only the alarm icon and alarm sound on the man-machine interaction system 105 are prompted to realize level 1 warning; when the drowsiness level reaches level 8 or 9, the icon flashing frequency is increased, the icon color is changed, the alarm sound volume is increased, the alarm time is increased, and even the steering wheel or seat vibration is increased, to realize level 2 warning with stronger warning effect.

[0067] In addition to determining the fatigue degree of the driver according to the driving performance, the domain controller 104 also learns in depth in real time through the driving habits of the driver. By taking the driving habits of the driver as a training set, based on a neural network algorithm, the corresponding relationship between the calibrated driving behavior and the fatigue level is fine-tuned by a certain proportion, to ensure that the performance of the system is more in line with the driving habits of the driver and more humanized. That is, the domain controller 104 adjusts the judgment of the fatigue degree of the driver by taking the data of the normal driving of the driver as a training set based on a neural network algorithm.

[0068] The fatigue level F of the driver is matched with the fatigue level on the KKS. The fatigue level F of the driver is defined as F = R1f + R2f' + Ω. F is used to match the range corresponding to each level of the fatigue level table, to trigger the fatigue warning signal of the corresponding level.

[0069] f' is the fatigue level calculated according to the learning and training of the data during normal driving, and its calculation method is the same as f; Ω is a set bias.

[0070] R1 and R2 are weight coefficients of f and f', respectively. R1 and R2 are obtained through a feedforward neural network algorithm, and R 1+ R2≤1, the adjustable range of R1 and R2 needs to be obtained through a large number of real vehicle tests in the functional development process. The value of R2 needs to be much smaller than R1, that is, the fatigue level calculated through the actual driving habits of the driver in the later stage can only be used as a comfort fine-tuning factor, and cannot have too great an impact on the actual performance of the system, causing too great a difference between the use effect and the fatigue level warning effect of the calibrated level in the research and development stage, so as to avoid affecting the type approval result of the compliance of relevant regulations in some countries or regions.

[0071] Considering that not all people have the same driving habits, for example, individual drivers may have lane deviation or abnormal steering wheel use even in non-fatigue situations, therefore, if the actual habits of each driver can be fine-tuned, the effect will be more personalized. However, because type approval in some countries or regions includes requirements for this function, the actual driving habits of each driver should not have a large impact on the overall fatigue level F, so as not to fail type approval. After all, KSS is also based on a large number of samples, and it is impossible to apply to every driver, therefore, the requirement R2 << R1 is to avoid the deviation of individual driving habits being too large, greatly affecting the product performance set during development. For example, set R1 = 95%, R2 = 5%, that is, the impact of individual driving habits on the actual performance of the product is only 5%.

[0072] The setting of fatigue level F in development needs to match KSS, since F is obtained by the joint influence of multiple factors, therefore, it cannot be guaranteed that adjusting the coefficients related to F can make the curve of F coincide with the curve of KSS, therefore, Ω is only used to cooperate with curve fitting, and the appropriate Ω needs to be calculated according to the deviation of F and KSS.

[0073] The corresponding relationship between the initial driving behavior of the driver and the fatigue level matched by calibration in the research and development stage can be used as the initial setting when the vehicle is delivered. After the vehicle is delivered, the user can freely choose to use this default mapping relationship, or choose to open the continuous learning function of the domain controller, and use the mapping relationship fine-tuned and continuously corrected by fusing the driving habits of the user.

[0074] Embodiment two

[0075] The embodiment provides a fatigue monitoring and early warning method, comprising the following steps:

[0076] S1, defining the range of the hand torque applied by the driver on the steering wheel in a normal driving situation; fitting the steering wheel rotation angle and steering wheel rotation angle velocity curve in normal driving, and setting a tolerance as the compliance interval of the steering wheel rotation angle and steering wheel rotation angle velocity in normal driving;

[0077] Determine the range corresponding to each level of the fatigue level reference table, and set the corresponding fatigue warning signal;

[0078] S2, the intelligent front camera detects road information, unexpected line pressing and lane center deviation driving behavior, and outputs the number X1 of unexpected line pressing and lane center deviation driving behavior of the vehicle in a specified period to the domain controller in real time;

[0079] The electric power steering system detects the hand torque applied by the driver on the steering wheel, the steering wheel rotation angle, the steering wheel rotation angular velocity, and outputs the time proportion P of the hand torque applied by the driver on the steering wheel exceeding the normal driving hand torque range in a specified period, and the time proportion Q of the steering wheel rotation angle and the steering wheel rotation angular velocity exceeding the compliance interval to the domain controller in real time;

[0080] The capacitive steering wheel detects the grip force applied by the driver on the steering wheel, judges whether the driver is off the steering wheel, and outputs the number X3 of times that the driver is off the steering wheel in a specified period to the domain controller in real time;

[0081] S3, define the initial driver fatigue level as f, then f=W i1 X1+W i2 X2+W i3 X3+θ;X2=K1P+K2Q;

[0082] Match f with the range corresponding to each level of the fatigue level table, and send the corresponding fatigue alarm signal instruction to the human-computer interaction system;

[0083] The domain controller also adjusts the judgment of the driver's fatigue level based on the neural network algorithm by taking the driver's operation habit as a training set;

[0084] Define the driver fatigue level as F, F=R1f+R2f’+Ω;

[0085] f’ is the fatigue level calculated according to the data during normal driving, and its calculation method is the same as f;

[0086] Match F with the range corresponding to each level of the fatigue level table, and send the corresponding fatigue alarm signal instruction to the human-computer interaction system;

[0087] S4, the human-computer interaction system sends a fatigue alarm signal after receiving the instruction of the domain controller.

[0088] Embodiment three

[0089] The embodiment provides a fatigue monitoring and early warning device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the fatigue monitoring and early warning method described above when executing the program.

[0090] Embodiment four

[0091] The embodiment provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to implement the fatigue monitoring and early warning method described above.

[0092] The above merely describes the preferred embodiments of the present application and is not used to limit the present application, and although the present application is described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or equivalently replace some technical features thereof. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A fatigue monitoring and alert system, characterized by: The domain controller, a human-computer interaction system in communication connection with the domain controller, a smart front camera, an electric power steering system, and a capacitive steering wheel are included. The smart front camera is used to detect vehicle driving information. The electric power steering system is used to detect state information of the steering wheel. The capacitive steering wheel is used to detect force on the steering wheel and determine whether the steering wheel is dropped. The domain controller is used to receive the state information and determine the fatigue degree of the driver based on the state information. The human-computer interaction system is used to receive instructions from the domain controller and send a fatigue warning signal. Define the initial fatigue level rating as f, then f = W i1 X1 + W i2 X2 + W i3 X3 + θ; W i1 , W i2 , W i3 are weight coefficients of X1, X2, X3, respectively; θ is a bias, which is obtained by fitting the actual vehicle test results and the theoretical results; X1 is the number of times of non-expected line pressing and lane center deviating driving behaviors of the vehicle in a period detected by the smart front camera; X2=K1P+K2Q; K1 and K2 are weight coefficients of P and Q, respectively; P is a time proportion of the hand torque on the steering wheel exceeding a preset range of hand torque in a period detected by the electric power steering system; Q is a time proportion of the steering wheel rotation angle and the steering wheel rotation angular velocity exceeding a compliance interval in a period; the compliance interval of the steering wheel rotation angle and the steering wheel rotation angular velocity is obtained based on a fitting curve of the steering wheel rotation angle and the steering wheel rotation angular velocity in normal driving and a tolerance setting; X3 is the number of times of hand dropping in a period; f is used to match a fatigue level table to determine the corresponding fatigue level.

2. The fatigue monitoring and alerting system of claim 1, wherein: The fatigue level table is a KSS drowsiness table.

3. The fatigue monitoring and alerting system of claim 1, wherein: The domain controller is further used to obtain data of the driver in normal driving as a training set, adjust the determination of the fatigue degree of the driver based on a neural network algorithm, and trigger a fatigue warning signal corresponding to the fatigue level.

4. The fatigue monitoring and alerting system of claim 3, wherein: The fatigue degree level of the driver is defined as F, F=R1f+R2f’+Ω; f’ is a fatigue degree level calculated according to the data in normal driving, and the calculation method is the same as that of f; Ω is a set bias; R1, R2 are weight coefficients of f, f', R1, R2 are obtained by a feedforward neural network algorithm, and R 1+ R2≤1, R2 R1; F is used to match the range corresponding to each level of the fatigue level table to trigger a fatigue warning signal of the corresponding level.

5. A fatigue monitoring and warning method applied to the fatigue monitoring and warning system according to claim 1, characterized in that, The method comprises the following steps: S1, the domain controller receives state information; S2, the domain controller determines the fatigue degree of the driver based on the state information; S3, the domain controller sends instructions to the human-computer interaction system to make the human-computer interaction system send a fatigue warning signal.

6. The fatigue monitoring and early warning method according to claim 5, wherein: the domain controller receives the number X1 of times of non-expected line pressing and lane center deviating driving behaviors of the vehicle in a specified period in real time; the domain controller receives a time proportion P of hand torque of the driver on the steering wheel exceeding a set normal driving hand torque range and a time proportion Q of the steering wheel rotation angle and the steering wheel rotation angular velocity exceeding a compliance interval in a specified period in real time; the domain controller receives the number X3 of times of hand dropping of the driver in a specified period in real time; Define the initial driver fatigue level as f, f = W i1 X1 + W i2 X2 + W i3 X3 + θ; X2 = K1P + K2Q; f is matched with the range corresponding to each level of the fatigue level table, and corresponding fatigue warning signal instructions are sent to the human-computer interaction system; X1 is obtained by detecting road information, non-expected line pressing, and lane center deviating driving behaviors by the smart front camera; P and Q are obtained by detecting hand torque of the driver on the steering wheel, the steering wheel rotation angle, and the steering wheel rotation angular velocity by the electric power steering system; and X3 is obtained by judging the grip force exerted on the steering wheel by the driver through the capacitive steering wheel.

7. The fatigue monitoring and alerting method of claim 6, wherein: The domain controller also adjusts the judgment of the driver's fatigue degree based on a neural network algorithm by taking the driver's operation habits as a training set; Define the driver fatigue degree level as F, F = R1f + R2f' + Ω; f' is the fatigue degree level calculated according to the data learned and trained during normal driving, and its calculation method is the same as f; Match F with the range corresponding to each level of the fatigue level table, and send the corresponding fatigue warning signal instruction to the human-computer interaction system.

8. A fatigue monitoring and warning device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that: The processor executes the program to implement the fatigue monitoring and warning method of any one of claims 5-7.

9. A computer readable storage medium having stored thereon a computer program, characterized in that: The program is executed by the processor to implement the fatigue monitoring and warning method of any one of claims 5-7.

Images