Method and apparatus for intelligent sensing and intervention of driver fatigue

Abstract

The application provides a method and device for intelligent perception and intervention of driving fatigue, wherein the method comprises the following steps: obtaining vehicle conditions by arranging sensors inside and outside the vehicle; obtaining basic data of the vehicle and driving behavior records of the driver; calculating a lane keeping rate of a vehicle trajectory according to the vehicle conditions; constructing a multi-factor coupling model of driving fatigue and calibrating model parameters according to the vehicle conditions, the basic data, the driving behavior records and the lane keeping rate; realizing intelligent perception of driving fatigue according to the multi-factor coupling model of driving fatigue, and realizing effective selection of a fatigue driving intervention mode through intervention and counterfactual analysis. The application provides an integrated solution for non-contact driving fatigue perception and intervention by coupling multi-factor effects, and can further serve other scenes such as driving interaction game perception.

Description

Technical Field

[0001] This invention belongs to the field of intelligent transportation. Background Technology

[0002] Fatigue driving and its derivative behaviors such as forced lane changes and speeding have a significant impact on road traffic safety. Intelligent perception and early warning of driver fatigue are of great significance for preventing accidents and reducing accident rates, and there is considerable research on this topic. Existing perception schemes mostly target driver's mental or muscular fatigue separately, using methods such as eye trackers, cerebral oxygenation monitoring, driver facial expressions, and steering wheel movements, as well as data fusion based on these data. However, because these methods may involve direct contact with the human body or have a monitoring function, the extent of their impact on driver fatigue remains unclear. Furthermore, because they neglect factors such as varying driver abilities, dynamic changes in the surrounding environment (visual environment, driving interaction), and their potential coupling effects, existing driver fatigue perception methods operate independently from the driver's and vehicle's perspectives, failing to establish a method and system for perceiving the coupled effects of driver, vehicle, and environmental factors. There are various intervention methods for driver fatigue, such as playing fast-paced music, cooling hands and neck, and using steering wheel vibration. Choosing the right intervention method at the right time, or a combination thereof, is an effective way to minimize the impact of driver fatigue on traffic safety. Based on this, a new automotive device system for preventing driver fatigue can be developed. Summary of the Invention

[0003] The present invention aims to at least partially solve one of the technical problems in the related art.

[0004] Therefore, the first objective of this invention is to propose a method for intelligent perception and intervention of driver fatigue, providing a feasible solution for non-contact driver fatigue perception.

[0005] The second objective of this invention is to provide a device for intelligent perception and intervention of driver fatigue.

[0006] The third objective of this invention is to provide a computer device.

[0007] The fourth objective of this invention is to provide a computer-readable storage medium.

[0008] To achieve the above objectives, a first aspect of the present invention provides a method for intelligent perception and intervention of driver fatigue, comprising: acquiring vehicle information, wherein the vehicle information is obtained by deploying sensors inside and outside the vehicle;

[0009] Acquire the vehicle's basic data and the driver's driving behavior records;

[0010] Calculate the lane keeping rate of the vehicle trajectory based on the vehicle information;

[0011] Based on the vehicle conditions, the basic data, the driving behavior records, and the lane keeping rate, a multi-factor coupled model of driving fatigue is constructed and the model parameters are calibrated.

[0012] Based on the aforementioned multi-factor coupling model of driving fatigue, intelligent perception of driving fatigue is achieved, and effective selection of fatigue driving intervention modes is realized through intervention and counterfactual analysis.

[0013] In addition, the intelligent driving fatigue perception method according to the above embodiments of the present invention may also have the following additional technical features:

[0014] Furthermore, in one embodiment of the present invention, the vehicle conditions include, but are not limited to, any one or more of the following:

[0015] External conditions, including road alignment, visibility, tree obstruction, weather conditions, and traffic flow;

[0016] Conditions inside the vehicle, including carbon dioxide concentration, temperature, and humidity;

[0017] The driver's personal circumstances, including vehicle operation behavior and mental state.

[0018] Furthermore, in one embodiment of the present invention, the lane keeping rate for calculating the vehicle trajectory includes, but is not limited to, any one or more of the following methods:

[0019] Determine directly based on the distance from the lane centerline / edge line;

[0020] The determination is made indirectly based on statistics of the distance from the lane centerline / edgeline, wherein the statistics of the edgeline distance include: mean, variance, and sample dynamics.

[0021] Furthermore, in one embodiment of the present invention, the construction of a multi-factor coupling model of driving fatigue and the calibration of model parameters, wherein the multi-factor coupling model includes:

[0022] Causal reasoning of driver fatigue is achieved through multi-factor coupling modeling and a model that can characterize the coupling effect of multi-factors. Intervention and counterfactual reasoning are realized on the calibrated multi-factor coupling model, thereby obtaining an effective selection of intervention mode.

[0023] The intervention modes include, but are not limited to, any one or more of the following:

[0024] Psychological stimulation, including playing fast-paced music and warning audio;

[0025] Sensory stimuli include cooling of the hands and wrists, and vibration of the steering wheel.

[0026] Furthermore, in one embodiment of the present invention, constructing a multi-factor coupled model of driving fatigue includes:

[0027] Define discrete and continuous variables to identify major influencing factors that cause driver fatigue, including the driver's initial state, the surrounding environment of the vehicle, and the situation inside the vehicle.

[0028] For cases where discrete and continuous variables coexist, the model is solved based on dynamic discretization.

[0029] For fatigue detected simultaneously in driving behavior records and lane keeping rate, the values ​​of the corresponding influencing factors are extracted for model calibration.

[0030] Furthermore, in one embodiment of the present invention, the content of the model calibration includes:

[0031] Distribution of discrete variables;

[0032] Distribution of continuous variables;

[0033] The relationships between the variables.

[0034] To achieve the above objectives, a second aspect of the present invention provides a device for intelligent perception and intervention of driver fatigue, comprising the following modules:

[0035] The acquisition module is used to acquire vehicle information, which is obtained by deploying sensors inside and outside the vehicle.

[0036] The recording module is used to acquire basic data of the vehicle and records of the driver's driving behavior;

[0037] The calculation module is used to calculate the lane keeping rate of the vehicle trajectory based on the vehicle conditions;

[0038] The calibration module is used to construct a multi-factor coupled model of driving fatigue based on the vehicle conditions, the basic data, the driving behavior records, and the lane keeping rate, and to calibrate the model parameters.

[0039] The perception module is used to achieve intelligent perception of driving fatigue based on the multi-factor coupling model of driving fatigue, and to achieve effective selection of fatigue driving intervention mode through intervention and counterfactual analysis.

[0040] Furthermore, in one embodiment of the present invention, the calibration module further includes:

[0041] The building block, used to construct a multi-factor coupled model of driver fatigue, further includes:

[0042] Define discrete and continuous variables to identify major influencing factors that cause driver fatigue, including the driver's initial state, the surrounding environment of the vehicle, and the situation inside the vehicle.

[0043] For cases where discrete and continuous variables coexist, the model is solved based on dynamic discretization.

[0044] For fatigue detected simultaneously in driving behavior records and lane keeping rate, the values ​​of the corresponding influencing factors are extracted for model calibration.

[0045] The content of the model calibration includes:

[0046] Distribution of discrete variables;

[0047] Distribution of continuous variables;

[0048] The relationships between the variables.

[0049] To achieve the above objectives, a third aspect of the present invention provides a computer device, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the method for intelligent perception and intervention of driving fatigue as described above.

[0050] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the method for intelligent perception and intervention of driving fatigue as described above.

[0051] The main advantages of the intelligent perception and intervention method and device for driver fatigue proposed in the embodiments of the present invention are: (1) Compared with measurement methods such as EEG and ECG, the non-contact working method of the present invention can evaluate the fatigue driving state by comprehensively considering the influence of the driver and the traffic flow around the vehicle while reducing the impact on driving operation; (2) The multi-factor coupling model involved in the present invention can also integrate the advantages and disadvantages of different intervention modes into the model, and realize the integrated system of driver fatigue and perception through reasoning, intervention and counterfactual reasoning; (3) The method involved in the present invention can further serve other scenarios such as risk perception in driving interaction game. Attached Figure Description

[0052] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0053] Figure 1 This is a flowchart illustrating a method for intelligent perception and intervention of driver fatigue provided in an embodiment of the present invention.

[0054] Figure 2 This is a schematic diagram of sensor layout provided in an embodiment of the present invention.

[0055] Figure 3 This is a schematic diagram of a questionnaire survey provided in an embodiment of the present invention.

[0056] Figure 4 This is a schematic diagram of driving fatigue recognition based on trajectory data provided in an embodiment of the present invention.

[0057] Figure 5 This is a schematic diagram of the multi-factor coupling model of driving fatigue constructed according to an embodiment of the present invention.

[0058] Figure 6 This is a schematic flowchart of a device for intelligent perception and intervention of driver fatigue provided in an embodiment of the present invention. Detailed Implementation

[0059] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0060] The following describes, with reference to the accompanying drawings, a method and apparatus for intelligent perception and intervention of driver fatigue according to embodiments of the present invention.

[0061] Example 1

[0062] Figure 1 This is a flowchart illustrating a method for intelligent perception and intervention of driver fatigue provided in an embodiment of the present invention.

[0063] like Figure 1 As shown, the method for intelligent perception and intervention of driver fatigue includes the following steps:

[0064] S101: Obtain vehicle status information, which is obtained by deploying sensors inside and outside the vehicle.

[0065] Furthermore, in one embodiment of the present invention, the vehicle conditions include, but are not limited to, any one or more of the following:

[0066] External conditions, including road alignment, visibility, tree obstruction, weather conditions, and traffic flow;

[0067] Conditions inside the vehicle, including carbon dioxide concentration, temperature, and humidity;

[0068] The driver's personal circumstances, including vehicle operation behavior and mental state.

[0069] The purpose of deploying sensors includes, but is not limited to, acquiring information on traffic flow around the vehicle, the surrounding line of sight, and the driver's operating behavior inside the vehicle.

[0070] 1) Traffic flow conditions: including but not limited to the lane-changing rate of vehicles in the lanes to the left and right of the lane where the sensing vehicle is located, and the average distance between vehicles in front of the lane where the sensing vehicle is located. Among them, the calculation method of lane-changing rate lc includes but is not limited to lc = (lane-changing vehicles / total number of vehicles passing).

[0071] 2) Environmental visibility: including but not limited to whether the vehicle is operating during the day or night, and whether there are weather events such as rain or fog that may affect it.

[0072] 3) Driver's operating behavior: including but not limited to the driver's grip on the steering wheel, vehicle braking, acceleration and other behaviors.

[0073] S102: Obtain basic vehicle data and driver's driving behavior records;

[0074] Methods for basic data research and driving behavior recording include, but are not limited to, questionnaires and interviews. Basic data refers to static data used for multi-factor coupled modeling, while driving behavior records are one of the data sources for validating the modeling results.

[0075] S103: Calculate the lane keeping rate of the vehicle trajectory based on the vehicle conditions;

[0076] Furthermore, in one embodiment of the present invention, the lane keeping rate for calculating the vehicle trajectory includes, but is not limited to, any one or more of the following methods:

[0077] Determine directly based on the distance from the lane centerline / edge line;

[0078] The determination is made indirectly based on statistics of the distance from the lane centerline / edgeline, wherein the statistics of the edgeline distance include: mean, variance, and sample dynamics.

[0079] S104: Construct a multi-factor coupled model of driving fatigue based on vehicle conditions, basic data, driving behavior records, and lane keeping rate, and calibrate the model parameters;

[0080] Furthermore, in one embodiment of the present invention, a multi-factor coupling model of driving fatigue is constructed and the model parameters are calibrated, wherein the multi-factor coupling model includes:

[0081] Causal reasoning of driver fatigue is achieved through multi-factor coupling modeling and a model that can characterize the coupling effect of multi-factors. Intervention and counterfactual reasoning are realized on the calibrated multi-factor coupling model, thereby obtaining an effective selection of intervention mode.

[0082] The intervention modes include, but are not limited to, any one or more of the following: psychological stimulation, including playing fast-paced music and warning audio; and physical stimulation, including cooling of the hands and wrists and vibration of the steering wheel.

[0083] Furthermore, in one embodiment of the present invention, a multi-factor coupled model of driving fatigue is constructed, including:

[0084] Define discrete and continuous variables to identify major influencing factors that cause driver fatigue, including the driver's initial state, the surrounding environment of the vehicle, and the situation inside the vehicle.

[0085] This influencing factor is the basis of the basic data survey. After updating the influencing factor, the model needs to be recalibrated.

[0086] For cases where discrete and continuous variables coexist, the model is solved based on dynamic discretization.

[0087] Suppose X is a continuous node in the network, and its state space is Ω. X The probability density equation is f X By discretizing i D =(X1,X2,...,X) N This maps continuous nodes X to discrete variables, with the domain of node X mapped to a hypercube. For example, R can be mapped to w. ( X 1) =(-∞,0],w ( X 2) = (0,1] and w ( X 3) = (0, +∞). Therefore, f X To be gathered Ψ X ={w x (l) The probability distribution P(X∈w) on} x (l) The purpose of dynamic discretization is to find an optimal set Ψ within the divided intervals. X And the optimal constant equation f, such that when the dynamic discretization process converges, f is approximately close to f X .

[0088] For fatigue detected simultaneously in driving behavior records and lane keeping rate, the values ​​of the corresponding influencing factors are extracted for model calibration.

[0089] Furthermore, in one embodiment of the present invention, the content of the model calibration includes:

[0090] Distribution of discrete variables; distribution of continuous variables; relationships between variables.

[0091] S105: Based on the multi-factor coupling model of driving fatigue, intelligent perception of driving fatigue is realized, and effective selection of fatigue driving intervention mode is achieved through intervention and counterfactual analysis.

[0092] After the model calibration is completed, the fatigue driving intelligent perception method and system proposed in this patent can be implemented using methods such as forward belief propagation and backward counterfactual intervention based on Bayesian causal networks. Since new influencing factors can be added with the support of offline system calibration, the perception method and system have good scalability.

[0093] Example 2

[0094] During the system offline calibration phase

[0095] The first step is the deployment of sensors for intelligent driver fatigue detection (see schematic diagram). Figure 2 The system comprises a sensor array, where sensors 1 and 3 monitor lane-changing rates on both sides, sensor 2 monitors the distance to the vehicle in front in the current lane, sensor 4 monitors driver grip strength, sensor 5 monitors the frequency of the driver's braking, and sensor 6 monitors the frequency of the driver's accelerator pedal presses. Sensor 4 must be capable of detecting pressure. Other sensors include, but are not limited to, cameras and millimeter-wave radar. Additionally, the vehicle is equipped with high-precision positioning for verification during system calibration. In practical applications, sensor deployment includes, but is not limited to, the types described above.

[0096] The second step involves a questionnaire survey and driving records for the driving test (example shown). Figure 3 (As shown), the survey is divided into two parts: one filled out by the driver before the test begins, and the other filled out by the experimenter during the test. The survey content includes, but is not limited to, the following: Figure 3 Examples are provided. The survey of drivers mainly includes their basic driving parameters and initial values ​​of driving status. Experimenters need to undergo standardized training and record the time when perceptible fatigue behavior, forced lane change behavior, and speeding behavior occur during the driving process.

[0097] The third step is to calculate the lane keeping rate of the vehicle trajectory, extract the time corresponding to fatigue behavior, and cross-validate it with the time corresponding to fatigue behavior recorded by the driver. For example... Figure 4 The diagram shows a trajectory segment. Assume the distance between the vehicle and the centerline of its lane is x, and the distance between the lane centerline and the lane edge is b. Methods for calculating fatigue time include, but are not limited to, the following criteria:

[0098]

[0099] If count ≥ n within period T, then driver fatigue is considered to have occurred; otherwise, it is considered not to have occurred. Extract the time information at the corresponding location.

[0100] The fourth step is to construct a multi-factor coupled model of driving fatigue and calibrate the model parameters. Establishing a model of multi-factor coupling includes, but is not limited to, Bayesian causal network models and other related methods. Taking a Bayesian causal network as an example, nodes can be defined as discrete or continuous variables, and the results can be solved based on a dynamic discretization method. Figure 5 As shown, driver fatigue is categorized into three main factors: initial driving state, surrounding conditions, and in-vehicle conditions. 1) Initial driving state encompasses basic information obtained from a questionnaire survey; 2) Surrounding conditions include lane-changing rates of the left or right lane and the average lane spacing ahead; 3) In-vehicle conditions include steering wheel grip strength and the frequency of pressing the accelerator and brake pedals. The modeling factors include, but are not limited to, the above-mentioned factors. Adding new factors requires a new process of calibrating the coupling factors.

[0101] After offline parameter calibration, a multi-factor coupling model is combined with sensors for online intelligent perception of driver fatigue. The appropriate intervention mode is selected through intervention and counterfactual analysis of the calibrated model. Specifically, the model's intervention analysis addresses the question, "What effect will be achieved if certain intervention measures are taken?"; the model's counterfactual analysis addresses the question, "What intervention measures are needed to achieve what goal?". Since driver fatigue is not caused by a single factor, and the degree of interaction between multiple factors is difficult to directly characterize, this invention provides a framework that can uniformly express various influencing factors and intervention modes, combined with an onboard sensor system to realize a practically deployed driver fatigue perception device.

[0102] To achieve the above embodiments, the present invention also proposes a device for intelligent perception and intervention of driver fatigue.

[0103] Figure 6 This is a schematic diagram of the structure of a device for intelligent perception and intervention of driver fatigue provided in an embodiment of the present invention.

[0104] like Figure 6 As shown, the intelligent driver fatigue perception and intervention device includes: an acquisition module 100, a recording module 200, a calculation module 300, a calibration module 400, and a perception module 500, wherein...

[0105] The acquisition module is used to acquire vehicle information, which is obtained by sensors deployed inside the vehicle.

[0106] The recording module is used to acquire basic vehicle data and driver driving behavior records;

[0107] The calculation module is used to calculate the lane keeping rate of the vehicle trajectory based on the vehicle conditions;

[0108] The calibration module is used to construct a multi-factor coupled model of driving fatigue based on vehicle conditions, basic data, driving behavior records, and lane keeping rate, and to calibrate the model parameters.

[0109] The perception module is used to achieve intelligent perception of driving fatigue based on the multi-factor coupling model of driving fatigue, and to achieve effective selection of fatigue driving intervention mode through intervention and counterfactual analysis.

[0110] Furthermore, in one embodiment of the present invention, the calibration module further includes:

[0111] The building block, used to construct a multi-factor coupled model of driver fatigue, further includes:

[0112] Define discrete and continuous variables to identify major influencing factors that cause driver fatigue, including the driver's initial state, the surrounding environment of the vehicle, and the situation inside the vehicle.

[0113] For cases where discrete and continuous variables coexist, the model is solved based on dynamic discretization.

[0114] For fatigue detected simultaneously in driving behavior records and lane keeping rate, the values ​​of the corresponding influencing factors are extracted for model calibration.

[0115] The model calibration includes the following:

[0116] Distribution of discrete variables;

[0117] Distribution of continuous variables;

[0118] The relationships between the variables.

[0119] To achieve the above objectives, a third aspect of the present invention provides a computer device, characterized in that it includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the method for intelligent perception and intervention of driving fatigue as described above.

[0120] To achieve the above objectives, a fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the method for intelligent perception and intervention of driving fatigue as described above.

[0121] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0122] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0123] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for intelligent perception and intervention of driver fatigue, characterized in that, Includes the following steps: Vehicle information is obtained by deploying sensors inside and outside the vehicle. Acquire the vehicle's basic data and the driver's driving behavior records; The lane keeping rate of the vehicle trajectory is calculated based on the vehicle conditions; the calculation of the lane keeping rate of the vehicle trajectory includes, but is not limited to, any one or more of the following methods: directly determining based on the distance from the lane centerline / edgeline; The determination is made indirectly based on statistics of the distance from the lane centerline / edgeline, wherein the statistics of the edgeline distance include: mean, variance, and sample dynamics; Based on the vehicle conditions, the basic data, the driving behavior records, and the lane keeping rate, a multi-factor coupled model of driving fatigue is constructed and the model parameters are calibrated. Based on the aforementioned multi-factor coupling model of driving fatigue, intelligent perception of driving fatigue is achieved, and effective selection of fatigue driving intervention modes is realized through intervention and counterfactual analysis. The construction of a multi-factor coupled model of driving fatigue and the calibration of model parameters, wherein the multi-factor coupled model includes: Causal reasoning of driver fatigue is achieved through multi-factor coupling modeling and a model that can characterize the coupling effect of multi-factors. Intervention and counterfactual reasoning are realized on the calibrated multi-factor coupling model, thereby obtaining an effective selection of intervention mode. The intervention modes include, but are not limited to, any one or more of the following: Psychological stimulation, including playing fast-paced music and warning audio; Sensory stimulation, including cooling of hands and wrists, and steering wheel vibration; The construction of the multi-factor coupled model of driving fatigue includes: Define discrete and continuous variables to identify major influencing factors that cause driver fatigue, including the driver's initial state, the surrounding environment of the vehicle, and the situation inside the vehicle. For cases where discrete and continuous variables coexist, the model is solved based on dynamic discretization. For fatigue detected simultaneously in driving behavior records and lane keeping rate, the values ​​of the corresponding influencing factors are extracted for model calibration. The content of the model calibration includes: Distribution of discrete variables; Distribution of continuous variables; The relationships between the variables.

2. The method according to claim 1, characterized in that, The vehicle situation includes, but is not limited to, any one or more of the following: External conditions, including road alignment, visibility, tree obstruction, weather conditions, and traffic flow; Conditions inside the vehicle, including carbon dioxide concentration, temperature, and humidity; The driver's personal circumstances, including vehicle operation behavior and mental state.

3. A device for intelligent perception and intervention of driver fatigue, characterized in that, Includes the following modules: The acquisition module is used to acquire vehicle information, which is obtained by deploying sensors inside and outside the vehicle. The recording module is used to acquire basic data of the vehicle and records of the driver's driving behavior; The calculation module is used to calculate the lane keeping rate of the vehicle trajectory based on the vehicle conditions; the calculation of the lane keeping rate of the vehicle trajectory includes, but is not limited to, any one or more of the following methods: directly determining based on the distance from the lane centerline / edge line; The determination is made indirectly based on statistics of the distance from the lane centerline / edgeline, wherein the statistics of the edgeline distance include: mean, variance, and sample dynamics; The calibration module is used to construct a multi-factor coupled model of driving fatigue based on the vehicle conditions, the basic data, the driving behavior records, and the lane keeping rate, and to calibrate the model parameters. The perception module is used to realize intelligent perception of driving fatigue based on the multi-factor coupling model of driving fatigue, and to realize effective selection of fatigue driving intervention mode through intervention and counterfactual analysis. The construction of a multi-factor coupled model of driving fatigue and the calibration of model parameters, wherein the multi-factor coupled model includes: Causal reasoning of driver fatigue is achieved through multi-factor coupling modeling and a model that can characterize the coupling effect of multi-factors. Intervention and counterfactual reasoning are realized on the calibrated multi-factor coupling model, thereby obtaining an effective selection of intervention mode. The intervention modes include, but are not limited to, any one or more of the following: Psychological stimulation, including playing fast-paced music and warning audio; Sensory stimulation, including cooling of hands and wrists, and steering wheel vibration; The calibration module further includes: The building block, used to construct a multi-factor coupled model of driver fatigue, further includes: Define discrete and continuous variables to identify major influencing factors that cause driver fatigue, including the driver's initial state, the surrounding environment of the vehicle, and the situation inside the vehicle. For cases where discrete and continuous variables coexist, the model is solved based on dynamic discretization. For fatigue detected simultaneously in driving behavior records and lane keeping rate, the values ​​of the corresponding influencing factors are extracted for model calibration. The content of the model calibration includes: Distribution of discrete variables; Distribution of continuous variables; The relationships between the variables.

4. A computer device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the method for intelligent perception and intervention of driving fatigue as described in any one of claims 1-2.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for intelligent perception and intervention of driving fatigue as described in any one of claims 1-2.

Images