A driver state multi-dimensional recognition system and device

Abstract

The application discloses a driver state multi-dimensional recognition system and equipment, and relates to the technical field of vehicle safety.The system comprises an initial individual baseline construction module, a real-time data acquisition module, a smooth driving period identification module, an individual baseline dynamic updating module and a driver state judgment module; the initial individual baseline construction module collects smooth data of the first use of the driver under multiple typical scenes, and generates a scene-specific personalized baseline; the real-time data acquisition module acquires real-time multi-dimensional data and scene characteristic data and identifies the scene; the smooth driving period identification module matches the scene baseline to screen smooth segments and combine the periods; the individual baseline dynamic updating module periodically corrects the baseline based on the smooth data; and the driver state judgment module calculates single-dimensional and multi-dimensional cooperative deviation values, and judges the state in stages.The application solves the problem that the group standard ignores individual differences, improves the recognition accuracy and scene adaptability, and guarantees driving safety.

Description

Technical Field

[0001] This invention relates to the field of vehicle safety technology, specifically to a multi-dimensional driver status recognition system and device. Background Technology

[0002] In the field of road traffic safety, accurately identifying driver status is crucial for preventing traffic accidents. Currently, most driver status recognition systems and devices on the market rely primarily on group standards to determine whether a driver is in an abnormal state. For example, in heart rate monitoring, the common criterion is a group average threshold such as "normal heart rate 60-100 beats / minute"; similarly, other dimensions such as blinking frequency and operating habits are mostly based on general group data.

[0003] However, this group-based identification method has significant flaws. Drivers exhibit substantial individual differences; drivers of different ages, professions, and physical conditions have vastly different physiological indicators and behavioral habits under normal conditions. For example, athletes, after long-term training, often have resting heart rates below 60 beats per minute, which could easily be misjudged as abnormal heart rates according to traditional group standards. Older drivers, due to declining physical function, have different normal reaction times and operational strengths compared to younger drivers. Professional drivers, due to long-term driving experience, have developed unique operating habits and fatigue tolerance levels.

[0004] Traditional systems, by neglecting individual differences, struggle to accurately identify the true state of individuals, leading to frequent misjudgments and missed detections. Misjudgments can trigger unnecessary interference and alarms, impacting the driving experience; missed detections fail to identify potential dangerous driver conditions in a timely manner, increasing the risk of traffic accidents. Therefore, existing driver state recognition technologies and equipment based on group standards are no longer sufficient to meet the growing demands for traffic safety. There is an urgent need for new technologies and equipment that can fully consider individual differences and achieve accurate driver state identification to improve road traffic safety. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-dimensional driver status recognition system and device to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] A multi-dimensional driver status recognition system includes: an initial individual baseline construction module, a real-time data acquisition module, a stable driving period recognition module, an individual baseline dynamic update module, and a driver status judgment module;

[0008] The initial individual baseline construction module is based on a preset baseline calibration process. It collects multi-dimensional data of drivers when they use the system for the first time in various typical driving scenarios. Based on the collected multi-dimensional data, it generates the initial individual baseline set of the driver in the corresponding scenario. The multi-dimensional data includes the driver's physiological characteristics, behavioral characteristics and operational characteristics in a stable driving state in each scenario. The initial individual baseline set includes the physiological baseline, behavioral baseline and operational baseline of each scenario.

[0009] The real-time data acquisition module collects real-time multi-dimensional data and real-time scene feature data of the driver during subsequent driving. The real-time multi-dimensional data includes the driver's real-time physiological feature data, real-time behavioral feature data and real-time operational feature data, while the real-time scene feature data is used to characterize the current driving scenario type.

[0010] The smooth driving period identification module performs hierarchical analysis on real-time multi-dimensional data and real-time scene feature data. First, it matches the corresponding scene baseline based on the real-time scene feature data, and then combines the scene baseline to filter smooth driving segments in the real-time multi-dimensional data. It merges continuous smooth driving segments into smooth driving periods. Among them, a smooth driving segment is a period in which the deviation between the real-time multi-dimensional data and the corresponding scene baseline is within a preset range and the driving state is stable.

[0011] The individual baseline dynamic update module periodically adjusts the corresponding baselines in the initial individual baseline set according to the physiological, behavioral, and operational dimensions based on real-time multi-dimensional data during the identified stable driving period, forming an updated individual baseline set. The periodic adjustment includes independent correction of the feature parameters of each dimension baseline and collaborative correction of cross-dimensional correlation parameters.

[0012] The driver status assessment module compares real-time multi-dimensional data with the updated individual baseline against the current scenario baseline, calculates the single-dimensional deviation value and multi-dimensional collaborative deviation value of each dimension of data, and determines whether the driver's current status is abnormal based on the combination of the single-dimensional deviation value and the multi-dimensional collaborative deviation value.

[0013] Furthermore, the initial individual baseline construction module includes a scene data acquisition unit, a baseline calculation unit, and a baseline set integration unit;

[0014] The scene data acquisition unit identifies a set of typical driving scenarios, guides the driver to complete smooth driving operations in each scenario, and synchronously collects multi-dimensional data sequences during the process through multi-dimensional sensors. The set of typical driving scenarios includes urban road scenarios, highway scenarios, nighttime road scenarios, and rain and snow weather scenarios. The multi-dimensional data sequences include physiological characteristic data sequences, behavioral characteristic data sequences, and operational characteristic data sequences. The physiological characteristic data sequences include continuously collected heart rate data and blink duration data. The behavioral characteristic data sequences include continuously collected eye gaze angle data and head rotation angle data. The operational characteristic data sequences include continuously collected steering wheel rotation angle data and accelerator pedal depth data.

[0015] The baseline calculation unit performs statistical analysis on multi-dimensional data sequences under typical driving scenarios to obtain the mean and standard deviation of each feature parameter, generating physiological baselines, behavioral baselines, and operational baselines for each scenario. The physiological baseline generated from the physiological feature data sequence is expressed as: Bp(s)={(μp1(s),σp1(s)),(μp2(s),σp2(s)),...,(μpn(s),σpn(s))}, where s represents a typical driving scenario, μpi(s) is the mean of the i-th physiological feature in scenario s, and σpi(s) is the standard deviation of the i-th physiological feature in scenario s, with i ranging from 1 to n. Similarly, the behavioral feature data sequence and operational feature data sequence are processed in the same way to generate the behavioral baseline Bb(s) and operational baseline Bo(s) for each scenario, and the expressions for the behavioral and operational baselines are consistent with those for the physiological baseline.

[0016] The baseline set integration unit integrates the physiological baseline, behavioral baseline, and operational baseline for each scenario to form an initial individual baseline set B0 containing multi-dimensional baselines corresponding to each scenario, where B0 = {Bp(s), Bb(s), Bo(s) | s ∈ typical driving scenario set}.

[0017] Furthermore, the real-time data acquisition module includes a scene environment parameter acquisition unit, a driving scene classification unit, and a multi-dimensional real-time data acquisition unit;

[0018] The scene environment parameter acquisition unit collects real-time scene feature data through scene recognition sensors. The real-time scene feature data includes road type parameters, light intensity parameters, and precipitation intensity parameters. The driving scene classification unit classifies and matches the real-time scene feature data acquired by the scene environment parameter acquisition unit, and identifies the current driving scene type based on the matching results. After the driving scene classification unit identifies the current driving scene type, the multi-dimensional real-time data acquisition unit simultaneously collects the driver's real-time physiological feature data, real-time behavioral feature data, and real-time operational feature data in that scene.

[0019] Furthermore, the driving scenario classification unit classifies and matches the real-time environmental parameters acquired by the scenario environment parameter acquisition unit, and identifies the current driving scenario type based on the matching results. The specific process is as follows:

[0020] A scenario classification rule base is pre-built, which includes the environmental parameter threshold ranges corresponding to each typical driving scenario (urban road scenario, highway scenario, night road scenario, rain and snow weather scenario). The road type parameter threshold range includes the road width range, lane number range, and speed limit range corresponding to urban roads. The light intensity parameter threshold range includes the light intensity range corresponding to night roads. The precipitation intensity parameter threshold range includes the precipitation intensity range corresponding to rain and snow weather.

[0021] The system receives real-time scene feature data output from the scene environment parameter acquisition unit. This real-time scene feature data includes real-time road type parameter R, real-time illumination intensity parameter L, and real-time precipitation intensity parameter W. The system then matches the real-time road type parameter R with the threshold range of road type parameters for each scene in the scene classification rule base to obtain a road type matching result; matches the real-time illumination intensity parameter L with the threshold range of illumination intensity parameters for each scene in the scene classification rule base to obtain an illumination intensity matching result; and matches the real-time precipitation intensity parameter W with the threshold range of precipitation intensity parameters for each scene in the scene classification rule base to obtain a precipitation intensity matching result.

[0022] Based on the above three matching results, the current driving scenario type is determined using a scenario matching degree calculation model. When all three matching results corresponding to a typical driving scenario meet the threshold requirements, or when two core matching results (road type matching result and light intensity matching result) are met and the third matching result is not conflicting, the current driving scenario type is determined to be that typical driving scenario. The scenario matching degree S(s) is calculated as follows: S(s) = α·S_R(s) + β·S_L(s) + γ·S_W(s), where s represents the typical driving scenario, S_R(s) is the matching degree between the real-time road type parameter R and the threshold range of the road type parameter in scenario s, and S_L(s) is the matching degree between the real-time light intensity parameter L and the light intensity in scenario s. The matching degree of the threshold range of the degree parameter; S_W(s) is the matching degree of the real-time precipitation intensity parameter W with the threshold range of the precipitation intensity parameter of scene s; α, β, and γ are the weight coefficients of the matching degree of road type, the matching degree of light intensity, and the matching degree of precipitation intensity, respectively, and α+β+γ=1, where α≥β>γ, α ranges from 0.4 to 0.6, β ranges from 0.3 to 0.4, and γ ranges from 0.1 to 0.2; when S(s)≥0.8, the current driving scene type is determined to be scene s; when there are multiple scenes s with S(s)≥0.8, the scene with the largest S(s) is selected and the road type parameter is further refined for S_R(s) corresponding to α, so that the scene with the largest S_R(s) is selected as the current driving scene type.

[0023] Furthermore, the smooth driving period recognition module includes a scene baseline matching unit, a multi-dimensional deviation calculation unit, a smooth driving segment filtering unit, and a smooth driving period merging unit;

[0024] The scenario baseline matching unit receives the current driving scenario type s_current output by the driving scenario classification unit, and calls the sub-scenario baseline group B(s_current) corresponding to the current driving scenario type from the initial individual baseline set B0 generated by the baseline set integration unit in the initial individual baseline construction module; wherein, the sub-scenario baseline group B(s_current) includes the physiological baseline Bp(s_current), behavioral baseline Bb(s_current), and operational baseline Bo(s_current) of the sub-scenario.

[0025] The multi-dimensional deviation calculation unit receives real-time multi-dimensional data output by the multi-dimensional real-time data acquisition unit, and calculates the real-time deviation of each dimension by combining the scene baseline group B (s_current) called by the scene baseline matching unit.

[0026] The smooth driving segment filtering unit filters smooth driving segments from real-time multi-dimensional data based on the real-time deviation of each dimension output by the multi-dimensional deviation calculation unit and the driving state fluctuation judgment conditions.

[0027] The smooth driving period merging unit merges multiple smooth driving segments output by the smooth driving segment filtering unit to form a complete smooth driving period.

[0028] Furthermore, the specific process for calculating the real-time deviation of each dimension in the multi-dimensional deviation calculation unit is as follows:

[0029] The physiological deviation Dp(t) between the real-time physiological characteristic number P(t) and the physiological baseline Bp(s_current) is calculated using the following formula: Dp(t) = (1 / n)∑ i∈[1,n] [|pi(t)-μ_pi(s_current)| / (k·σ_pi(s_current))], where pi(t) is the real-time value of the i-th physiological feature at time t, μ_pi(s_current) is the baseline mean of the i-th physiological feature in the s_current scenario, σ_pi(s_current) is the baseline standard deviation of the i-th physiological feature in the s_current scenario, and k is a preset deviation correction coefficient; using the same calculation logic and formula as the above physiological deviation, the behavioral deviation Db(t) between the real-time behavioral feature data B(t) and the behavioral baseline Bb(s_current), and the operational deviation Do(t) between the real-time operational feature data O(t) and the operational baseline Bo(s_current) are calculated respectively;

[0030] The specific process for filtering smooth driving segments from real-time multi-dimensional data in the smooth driving segment filtering unit is as follows:

[0031] A preset stability deviation threshold D and a driving state fluctuation threshold Z are defined. For each time stamp t, if Dp(t)≤D, Db(t)≤D and Do(t)≤D, and the variance Var(O(t),O(t-1),O(t-2)) of the real-time operation feature data O(t) within three consecutive time stamps is ≤Z, then time t is determined to be a stable driving point. Stable driving points with consecutive time stamps are grouped into continuous data segments. When the duration of a continuous data segment is greater than or equal to a preset segment duration threshold, the continuous data segment is marked as a stable driving segment.

[0032] The specific analysis process for merging units during stable driving periods is as follows:

[0033] The preset segment interval threshold ε and the shortest duration threshold θ are used. If the time interval between two adjacent smooth driving segments is less than or equal to the segment interval threshold ε, the two segments are merged into a new smooth driving segment. The above merging operation is repeated until the interval between all adjacent segments is greater than the segment interval threshold ε. Segments with a duration greater than or equal to the shortest duration threshold θ after merging are selected and marked as the final smooth driving period.

[0034] Furthermore, the individual baseline dynamic update module includes a stationary data extraction unit, a single-dimensional baseline independent correction unit, a cross-dimensional correlation collaborative correction unit, and an updated baseline set integration unit.

[0035] The stable data extraction unit receives the stable driving period output by the stable driving period identification module and the current driving scenario type s_current output by the driving scenario classification unit, and extracts the multi-dimensional stable data sequence of the current driving scenario type within the stable driving period from the real-time multi-dimensional data. The multi-dimensional stable data sequence includes the physiological feature stable subsequence P_stable, the behavioral feature stable subsequence B_stable, and the operational feature stable subsequence O_stable.

[0036] The single-dimensional baseline independent correction unit, based on the multi-dimensional stationary data sequence output by the stationary data extraction unit, independently corrects the baselines of the corresponding scenarios in the initial individual baseline set B_0 according to physiological, behavioral, and operational dimensions. The specific analysis process is as follows:

[0037] A baseline update period T and a correction coefficient λ are preset, where 0 < λ < 1. When the system runtime reaches the update period T, an independent correction process is triggered to perform statistical analysis on the stationary subsequence P_stable of the physiological characteristics, calculating the stationary mean μp_new(s_current) and stationary standard deviation σp_new(s_current) of the i-th physiological characteristic. Based on the initial physiological baseline Bp_old(s_current) = {(μp_old(s_current), σp_old(s_current))}, the corrected physiological baseline parameters are calculated, where the corrected mean μp - update(s_current) is the standard deviation of the physiological characteristic. The formulas for calculating μp-update(s_current) and standard deviation σp-update(s_current) are: μp-update(s_current) = (1-λ)·μp_old(s_current) + λ·μp_new(s_current), σp-update(s_current) = (1-λ)·σp_old(s_current) + λ·σp_new(s_current); thus, the corrected physiological baseline is: Bp_update(s_current) = {(μp-update,σp-update(s_current))};

[0038] Using the same logic and formula as the above-mentioned independent correction of physiological baselines, the stationary subsequences of behavioral features B_stable and operational features O_stable are processed respectively to obtain the corrected behavioral baseline Bb_update(s_current) and operational baseline Bo_update(s_current).

[0039] The cross-dimensional correlation collaborative correction unit adjusts the baseline after independent correction in one dimension based on the changes in the correlation between multi-dimensional stationary data sequences; the specific analysis process is as follows:

[0040] The Pearson correlation coefficient method was used to calculate the physiological-behavioral correlation coefficient r_p-b, the physiological-operational correlation coefficient r_p-o, and the behavioral-operational correlation coefficient r_b-o, respectively. Historical cross-dimensional correlation coefficients r_old for the corresponding scenario s_current in the initial individual baseline set B0 were extracted, including r_p-b-old, r_p-o-old, and r_b-o-old. The corresponding change in correlation coefficient Δr = |r_new - r_old| was calculated, where r_new is the new correlation coefficient calculated based on stationary data. A preset correlation change threshold Δr0 was used. If Δr ≥ Δr0, collaborative correction was triggered; if Δr < Δr0, the correlation was determined to have no significant change, no collaborative correction was needed, and the baseline parameters after single-dimensional independent correction were retained.

[0041] The updated baseline set integration unit integrates the baselines of each dimension after independent correction in one dimension or collaborative correction across dimensions, forming the updated baseline group B_update(s_current) for the corresponding scenario, represented as:

[0042] B_update(s_current)={Bp_final(s_current),Bb_final(s_current),Bo_final(s_current)}, where Bp_final is the physiological baseline after independent or collaborative correction, Bb_final is the behavioral baseline after independent or collaborative correction, and Bo_final is the operational baseline after independent or collaborative correction; the updated baseline groups of all scenarios are integrated according to the structure of the initial individual baseline set B0 to form an updated individual baseline set Bnew covering all typical driving scenarios, and Bnew={B_update(s|s∈typical driving scenario set)}.

[0043] Furthermore, the driver status judgment module includes a scenario baseline call unit, a single-dimensional deviation value calculation unit, a multi-dimensional collaborative deviation value calculation unit, an abnormal status judgment unit, and a judgment result output unit;

[0044] The scenario baseline invocation unit receives the current driving scenario type s_current output by the driving scenario classification unit and the updated individual baseline set Bnew output by the individual baseline dynamic update module. It then calls the target scenario baseline group B_target(s_current) that matches s_current from the updated individual baseline set Bnew. The target scenario baseline group includes the updated physiological baseline Bp_final(s_current), behavioral baseline Bb_final(s_current), and operational baseline Bo_final(s_current).

[0045] The single-dimensional deviation calculation unit receives real-time multi-dimensional data output from the multi-dimensional real-time data acquisition unit and the target scene baseline group B_target(s_current) output from the scene baseline call unit, and calculates the single-dimensional deviation value of each feature according to the dimension. The specific analysis process is as follows:

[0046] With a preset deviation correction coefficient k1, and referring to the calculation formula of the real-time deviation in the multi-dimensional deviation calculation unit, calculate the single-dimensional deviation value Dp_single(t) of the physiological dimension, the single-dimensional deviation value Db_single(t) of the behavioral dimension, and the single-dimensional deviation value Do_single(t) of the operational dimension.

[0047] The multi-dimensional collaborative deviation calculation unit calculates the multi-dimensional collaborative deviation value based on the single-dimensional deviation value and the scenario-based weights of each dimension. It also incorporates the correction results of the inter-dimensional collaborative coefficients. The specific analysis is as follows:

[0048] A pre-defined scenario-based dimension weight library is used. For the current scenario s_current, the corresponding weights w_p(s_current), w_b(s_current), and w_o(s_current) are called, satisfying w_p(s_current) + w_b(s_current) + w_o(s_current) = 1, and the weight allocation matches the characteristics of the scenario. The initial multi-dimensional collaborative deviation value Dc_init(t) is calculated using the formula: Dc_init(t) = w_p(s_current)·Dp_single(t) + w_b(s_current)·Db_single(t) + w_o(s_current) ent)·Do_single(t); Based on the preset co-analysis time window T0, the synchronicity of the real-time single-dimensional deviation value is calculated using the Pearson correlation coefficient, thereby obtaining the physiological-operational co-coefficient Cp-o(t), the physiological-behavioral co-coefficient Cp-b(t), and the behavioral-operational co-coefficient Cb-o(t). The average of the three is taken as the final co-coefficient C(t); If C(t)≥0.6, then the coefficient of Dc_init(t) is corrected, and the corresponding formula is: Dc_final(t)=Dc_init(t)·(1+f·C(t)), where f represents the coefficient; If C(t)<0.6, then Dc_final(t)=Dc_init(t);

[0049] The abnormal state determination unit determines the driver's current state based on the single-dimensional deviation value and the corrected multi-dimensional collaborative deviation value, combined with a preset threshold. If Dp_single(t) < D1, Db_single(t) < D1, Do_single(t) < D1 and Dc_final(t) < D2, it is determined to be a "normal state", where D1 is the single-dimensional abnormal threshold and D2 is the multi-dimensional abnormal threshold.

[0050] If any single-dimensional deviation value is greater than or equal to D1 and Dc_final(t) < D2, it is judged as a "mildly abnormal state";

[0051] If Dc_final(t)≥D2 and C(t)<D3, it is determined to be a “moderately abnormal state”, where D3 is the synergistic enhancement threshold;

[0052] If Dc_final(t)≥D2 and C(t)≥D3, it is determined to be a "severe abnormal state";

[0053] The judgment result output unit receives the status judgment result output by the abnormal status judgment unit and outputs the judgment result in a preset format, which includes "current status type (normal / mild abnormal / moderate abnormal / severe abnormal), abnormal dimension identifier (such as "physiological dimension - heart rate abnormality"), single dimension deviation value details, multi-dimensional collaborative deviation value and collaborative coefficient".

[0054] A driver status multi-dimensional recognition device, including a multi-dimensional sensor group and a system early warning module;

[0055] The multi-dimensional sensor group includes physiological feature sensors, behavioral feature sensors, operational feature sensors, and scene recognition sensors, such as DMS cameras; the system warning module receives the judgment results output by the driver status judgment module and triggers corresponding warning interventions according to different abnormal status types. The warning interventions include light prompts, seat vibration, and voice warnings.

[0056] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention, through an initial individual baseline construction module, collects driver-specific stable data under various typical driving scenarios to generate scenario-specific personalized baselines. Combined with the periodic adjustments (including cross-dimensional collaborative correction) of the individual baseline dynamic update module, the baseline continuously adapts to changes in individual driver characteristics, fundamentally replacing group standards and significantly improving the accuracy of state recognition for special groups, reducing the risk of misjudgment and missed judgment. This invention accurately identifies the current scenario through a driving scenario classification unit. Modules such as stable driving period recognition and driver state judgment all call individual baselines matching the scenario. Furthermore, the multi-dimensional collaborative deviation value calculation introduces scenario-based weights, making the recognition logic fit the scenario characteristics, avoiding deviations caused by insufficient scenario adaptation, and maintaining stable recognition performance in complex environments. This invention, through the individual baseline dynamic update module, performs single-dimensional independent correction and cross-dimensional collaborative correction on the baseline based on stable driving data, balancing the weights of historical baselines and new data, capturing dimensional correlation changes, ensuring that the baseline always adapts to the driver's current state, solving the problem of "outdated and ineffective" traditional static baselines, and ensuring the recognition timeliness and stability of the system in long-term use. This invention categorizes driver status into four levels: "normal, mildly abnormal, moderately abnormal, and severely abnormal," labeling the abnormality dimension and coordination coefficient. The system can trigger differentiated warnings based on the classification results (such as light prompts for mild abnormalities and voice and vibration warnings for severe abnormalities). This avoids minor abnormalities from interfering with driving and strengthens intervention for high-risk states, thereby improving the driving experience while ensuring traffic safety. Attached Figure Description

[0057] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0058] Figure 1This is a schematic diagram of a multi-dimensional driver status recognition system according to the present invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] Please see Figure 1 The present invention provides the following technical solution:

[0061] A multi-dimensional driver status recognition system includes: an initial individual baseline construction module, a real-time data acquisition module, a smooth driving period recognition module, an individual baseline dynamic update module, and a driver status judgment module;

[0062] The initial individual baseline construction module is based on a preset baseline calibration process. It collects multi-dimensional data of drivers when they use the system for the first time in various typical driving scenarios. Based on the collected multi-dimensional data, it generates the initial individual baseline set of the driver in the corresponding scenario. The multi-dimensional data includes the driver's physiological characteristics, behavioral characteristics and operational characteristics in a stable driving state in each scenario. The initial individual baseline set includes the physiological baseline, behavioral baseline and operational baseline of each scenario.

[0063] The real-time data acquisition module collects real-time multi-dimensional data and real-time scene feature data of the driver during subsequent driving. The real-time multi-dimensional data includes the driver's real-time physiological feature data, real-time behavioral feature data and real-time operational feature data, while the real-time scene feature data is used to characterize the current driving scenario type.

[0064] The smooth driving period identification module performs hierarchical analysis on real-time multi-dimensional data and real-time scene feature data. First, it matches the corresponding scene baseline based on the real-time scene feature data, and then combines the scene baseline to filter smooth driving segments in the real-time multi-dimensional data. It merges continuous smooth driving segments into smooth driving periods. Among them, a smooth driving segment is a period in which the deviation between the real-time multi-dimensional data and the corresponding scene baseline is within a preset range and the driving state is stable.

[0065] The individual baseline dynamic update module periodically adjusts the corresponding baselines in the initial individual baseline set according to the physiological, behavioral, and operational dimensions based on real-time multi-dimensional data during the identified stable driving period, forming an updated individual baseline set. The periodic adjustment includes independent correction of the feature parameters of each dimension baseline and collaborative correction of cross-dimensional correlation parameters.

[0066] The driver status assessment module compares real-time multi-dimensional data with the updated individual baseline against the current scenario baseline, calculates the single-dimensional deviation value and multi-dimensional collaborative deviation value of each dimension of data, and determines whether the driver's current status is abnormal based on the combination of the single-dimensional deviation value and the multi-dimensional collaborative deviation value.

[0067] The initial individual baseline construction module includes a scene data acquisition unit, a baseline calculation unit, and a baseline set integration unit;

[0068] The scene data acquisition unit identifies a set of typical driving scenarios, guides the driver to complete smooth driving operations in each scenario, and synchronously collects multi-dimensional data sequences during the process through multi-dimensional sensors. The set of typical driving scenarios includes urban road scenarios, highway scenarios, nighttime road scenarios, and rain and snow weather scenarios. The multi-dimensional data sequences include physiological characteristic data sequences, behavioral characteristic data sequences, and operational characteristic data sequences. The physiological characteristic data sequences include continuously collected heart rate data and blink duration data. The behavioral characteristic data sequences include continuously collected eye gaze angle data and head rotation angle data. The operational characteristic data sequences include continuously collected steering wheel rotation angle data and accelerator pedal depth data.

[0069] The baseline calculation unit performs statistical analysis on multi-dimensional data sequences under typical driving scenarios to obtain the mean and standard deviation of each feature parameter, generating physiological baselines, behavioral baselines, and operational baselines for each scenario. The physiological baseline generated from the physiological feature data sequence is expressed as: Bp(s)={(μp1(s),σp1(s)),(μp2(s),σp2(s)),...,(μpn(s),σpn(s))}, where s represents a typical driving scenario, μpi(s) is the mean of the i-th physiological feature in scenario s, and σpi(s) is the standard deviation of the i-th physiological feature in scenario s, with i ranging from 1 to n. Similarly, the behavioral feature data sequence and operational feature data sequence are processed in the same way to generate the behavioral baseline Bb(s) and operational baseline Bo(s) for each scenario, and the expressions for the behavioral and operational baselines are consistent with those for the physiological baseline.

[0070] The baseline set integration unit integrates the physiological baseline, behavioral baseline, and operational baseline for each scenario to form an initial individual baseline set B0 containing multi-dimensional baselines corresponding to each scenario, where B0 = {Bp(s), Bb(s), Bo(s) | s ∈ typical driving scenario set}.

[0071] The real-time data acquisition module includes a scene environment parameter acquisition unit, a driving scene classification unit, and a multi-dimensional real-time data acquisition unit;

[0072] The scene environment parameter acquisition unit collects real-time scene feature data through scene recognition sensors. The real-time scene feature data includes road type parameters, light intensity parameters, and precipitation intensity parameters. The driving scene classification unit classifies and matches the real-time scene feature data acquired by the scene environment parameter acquisition unit, and identifies the current driving scene type based on the matching results. After the driving scene classification unit identifies the current driving scene type, the multi-dimensional real-time data acquisition unit simultaneously collects the driver's real-time physiological feature data, real-time behavioral feature data, and real-time operational feature data in that scene.

[0073] The driving scenario classification unit classifies and matches the real-time environmental parameters acquired by the scenario environment parameter acquisition unit, and identifies the current driving scenario type based on the matching results. The specific process is as follows:

[0074] A scenario classification rule base is pre-built, which includes the environmental parameter threshold ranges corresponding to each typical driving scenario (urban road scenario, highway scenario, night road scenario, rain and snow weather scenario). The road type parameter threshold range includes the road width range, lane number range, and speed limit range corresponding to urban roads. The light intensity parameter threshold range includes the light intensity range corresponding to night roads. The precipitation intensity parameter threshold range includes the precipitation intensity range corresponding to rain and snow weather.

[0075] The system receives real-time scene feature data output from the scene environment parameter acquisition unit. This real-time scene feature data includes real-time road type parameter R, real-time illumination intensity parameter L, and real-time precipitation intensity parameter W. The system then matches the real-time road type parameter R with the threshold range of road type parameters for each scene in the scene classification rule base to obtain a road type matching result; matches the real-time illumination intensity parameter L with the threshold range of illumination intensity parameters for each scene in the scene classification rule base to obtain an illumination intensity matching result; and matches the real-time precipitation intensity parameter W with the threshold range of precipitation intensity parameters for each scene in the scene classification rule base to obtain a precipitation intensity matching result.

[0076] Based on the above three matching results, the current driving scenario type is determined using a scenario matching degree calculation model. When all three matching results corresponding to a typical driving scenario meet the threshold requirements, or when two core matching results (road type matching result and light intensity matching result) are met and the third matching result is not conflicting, the current driving scenario type is determined to be that typical driving scenario. The scenario matching degree S(s) is calculated as follows: S(s) = α·S_R(s) + β·S_L(s) + γ·S_W(s), where s represents the typical driving scenario, S_R(s) is the matching degree between the real-time road type parameter R and the threshold range of the road type parameter in scenario s, and S_L(s) is the matching degree between the real-time light intensity parameter L and the light intensity in scenario s. The matching degree of the threshold range of the degree parameter; S_W(s) is the matching degree of the real-time precipitation intensity parameter W with the threshold range of the precipitation intensity parameter of scene s; α, β, and γ are the weight coefficients of the matching degree of road type, the matching degree of light intensity, and the matching degree of precipitation intensity, respectively, and α+β+γ=1, where α≥β>γ, α ranges from 0.4 to 0.6, β ranges from 0.3 to 0.4, and γ ranges from 0.1 to 0.2; when S(s)≥0.8, the current driving scene type is determined to be scene s; when there are multiple scenes s with S(s)≥0.8, the scene with the largest S(s) is selected and the road type parameter is further refined for S_R(s) corresponding to α, so that the scene with the largest S_R(s) is selected as the current driving scene type.

[0077] In this embodiment, the specific calculation process of the matching degree is as follows, taking S_R(s) as an example:

[0078] When R is within the threshold range of the road type parameter in scenario s, S_R(s) = 1. When at least two sub-parameters in R are within the threshold range of the road type parameter in scenario s, and the deviation rate of the remaining sub-parameters from the threshold is ≤20%, S_R(s) takes the value of 0.6-0.9. When only one sub-parameter in R is within the threshold range of the road type parameter in scenario s, or when there is a sub-parameter with a deviation rate >20%, S_R(s) takes the value of 0.1-0.5. When all sub-parameters in R are not within the threshold range of the road type parameter in scenario s, and the deviation rate of any sub-parameter from the threshold is >30%, S_R(s) = 0. Wherein, deviation rate = |real-time sub-parameter value - threshold boundary value| / threshold boundary value. If the real-time sub-parameter value is higher than the upper limit of the threshold, the upper limit of the threshold is taken as the threshold boundary value. If the real-time sub-parameter value is lower than the lower limit of the threshold, the lower limit of the threshold is taken as the threshold boundary value.

[0079] In this embodiment, it is assumed that the scene classification rule base contains nighttime urban roads s1 and nighttime highways s2, and the corresponding threshold ranges for road type parameters, light intensity parameters, and precipitation intensity parameters are as follows:

[0080] S1: Road width 15-25m, number of lanes 2-4, speed limit 40-60km / h; light intensity ≤200lux (night); precipitation intensity ≤5mm / h (no rain or snow);

[0081] S2: Road width 28-40m, number of lanes 4-6, speed limit 80-120km / h; light intensity ≤200lux (night); precipitation intensity ≤5mm / h (no rain or snow);

[0082] Assume the corresponding weighting coefficients are: α = 0.5 (road type), β = 0.3 (light intensity), γ = 0.2 (precipitation intensity), and satisfy α + β + γ = 1.

[0083] The real-time environmental parameters collected when a driver drives through an urban expressway at night are as follows:

[0084] Real-time road type parameter R: Road width 26m, number of lanes 4, speed limit 70km / h (between the thresholds of urban roads and highways).

[0085] Real-time light intensity parameter L: 150 lux (meets the "night" threshold ≤ 200 lux);

[0086] Real-time precipitation intensity parameter W: 0 mm / h (no rain or snow, meeting the requirement of "precipitation intensity ≤ 5 mm / h");

[0087] According to the matching degree formula S(s)=α·S_R(s)+β·S_L(s)+γ·S_W(s), the matching degree of s1 (nighttime urban roads) and s2 (nighttime highways) are calculated respectively:

[0088] The matching degree S(s1) for nighttime urban road s1:

[0089] S_R(s1) (Road type matching degree): The real-time road width of 26m slightly exceeds the s1 threshold (15-25m), but the number of lanes of 4 and the speed limit of 70km / h partially meet the s1 characteristics, so S_R(s1) is judged to be 0.8 (not a complete match but close to the threshold).

[0090] S_L(s1) (light intensity matching degree): 150 lux ≤ 200 lux, perfect match, S_L(s1) = 1;

[0091] S_W(s1) (Precipitation intensity matching degree): 0mm / h≤5mm / h, perfect match, S_W(s1)=1;

[0092] S(s1)=0.5×0.8+0.3×1+0.2×1=0.4+0.3+0.2=0.9≥0.8;

[0093] Similarly, the matching degree S(s2) of the highway s2 at night is calculated in the same way, and we get:

[0094] S(s2)=0.5×0.8+0.3×1+0.2×1=0.4+0.3+0.2=0.9≥0.8;

[0095] At this point, the matching degree of both nighttime urban road s1 and nighttime highway s2 is 0.9, satisfying the judgment condition of S(s)≥0.8, i.e., the situation of "multiple scenarios meeting the matching degree standard" occurs. The essence of this example is that the road parameters of urban expressways are between those of urban roads and highways, while the light and precipitation parameters of "no rain or snow at night" simultaneously meet the threshold requirements of both scenarios, ultimately resulting in the comprehensive matching degree of both scenarios reaching the threshold. According to the previously set "conflict handling rule" (when there are multiple scenarios s with S(s)≥0.8, select the scenario with the largest S(s) and further refine the road type parameters of S_R(s) corresponding to α, thereby selecting the scenario with the larger S_R(s) as the current driving scenario type), if the road type parameters are further refined in the future (such as adding sub-parameters such as "shoulder width" and "entrance / exit density"), the distinguishability of S_R(s) can be improved. Assume that a unique scenario is finally determined (such as urban expressways being classified as the "nighttime urban road" subclass).

[0096] The smooth driving period recognition module includes a scene baseline matching unit, a multi-dimensional deviation calculation unit, a smooth driving segment filtering unit, and a smooth driving period merging unit.

[0097] The scenario baseline matching unit receives the current driving scenario type s_current output by the driving scenario classification unit, and calls the sub-scenario baseline group B(s_current) corresponding to the current driving scenario type from the initial individual baseline set B0 generated by the baseline set integration unit in the initial individual baseline construction module; wherein, the sub-scenario baseline group B(s_current) includes the physiological baseline Bp(s_current), behavioral baseline Bb(s_current), and operational baseline Bo(s_current) of the sub-scenario.

[0098] The multi-dimensional deviation calculation unit receives real-time multi-dimensional data output by the multi-dimensional real-time data acquisition unit, and calculates the real-time deviation of each dimension by combining the scene baseline group B (s_current) called by the scene baseline matching unit.

[0099] The smooth driving segment filtering unit filters smooth driving segments from real-time multi-dimensional data based on the real-time deviation of each dimension output by the multi-dimensional deviation calculation unit and the driving state fluctuation judgment conditions.

[0100] The smooth driving period merging unit merges multiple smooth driving segments output by the smooth driving segment filtering unit to form a complete smooth driving period.

[0101] The specific process for calculating the real-time deviation of each dimension in the multi-dimensional deviation calculation unit is as follows:

[0102] The physiological deviation Dp(t) between the real-time physiological characteristic number P(t) and the physiological baseline Bp(s_current) is calculated using the following formula: Dp(t) = (1 / n)∑ i∈[1,n] [|pi(t)-μ_pi(s_current)| / (k·σ_pi(s_current))], where pi(t) is the real-time value of the i-th physiological feature at time t, μ_pi(s_current) is the baseline mean of the i-th physiological feature in the s_current scenario, σ_pi(s_current) is the baseline standard deviation of the i-th physiological feature in the s_current scenario, and k is a preset deviation correction coefficient; using the same calculation logic and formula as the above physiological deviation, the behavioral deviation Db(t) between the real-time behavioral feature data B(t) and the behavioral baseline Bb(s_current), and the operational deviation Do(t) between the real-time operational feature data O(t) and the operational baseline Bo(s_current) are calculated respectively;

[0103] The specific process for filtering smooth driving segments from real-time multi-dimensional data in the smooth driving segment filtering unit is as follows:

[0104] A preset stability deviation threshold D and a driving state fluctuation threshold Z are defined. For each time stamp t, if Dp(t)≤D, Db(t)≤D and Do(t)≤D, and the variance Var(O(t),O(t-1),O(t-2)) of the real-time operation feature data O(t) within three consecutive time stamps is ≤Z, then time t is determined to be a stable driving point. Stable driving points with consecutive time stamps are grouped into continuous data segments. When the duration of a continuous data segment is greater than or equal to a preset segment duration threshold, the continuous data segment is marked as a stable driving segment.

[0105] The specific analysis process for merging units during stable driving periods is as follows:

[0106] The preset segment interval threshold ε and the shortest duration threshold θ are used. If the time interval between two adjacent smooth driving segments is less than or equal to the segment interval threshold ε, the two segments are merged into a new smooth driving segment. The above merging operation is repeated until the interval between all adjacent segments is greater than the segment interval threshold ε. Segments with a duration greater than or equal to the shortest duration threshold θ after merging are selected and marked as the final smooth driving period.

[0107] The individual baseline dynamic update module includes a stationary data extraction unit, a single-dimensional baseline independent correction unit, a cross-dimensional correlation collaborative correction unit, and an updated baseline set integration unit.

[0108] The stable data extraction unit receives the stable driving period output by the stable driving period identification module and the current driving scenario type s_current output by the driving scenario classification unit, and extracts the multi-dimensional stable data sequence of the current driving scenario type within the stable driving period from the real-time multi-dimensional data. The multi-dimensional stable data sequence includes the physiological feature stable subsequence P_stable, the behavioral feature stable subsequence B_stable, and the operational feature stable subsequence O_stable.

[0109] The single-dimensional baseline independent correction unit, based on the multi-dimensional stationary data sequence output by the stationary data extraction unit, independently corrects the baselines of the corresponding scenarios in the initial individual baseline set B_0 according to physiological, behavioral, and operational dimensions. The specific analysis process is as follows:

[0110] A baseline update period T and a correction coefficient λ are preset, where 0 < λ < 1. When the system runtime reaches the update period T, an independent correction process is triggered to perform statistical analysis on the stationary subsequence P_stable of the physiological characteristics, calculating the stationary mean μp_new(s_current) and stationary standard deviation σp_new(s_current) of the i-th physiological characteristic. Based on the initial physiological baseline Bp_old(s_current) = {(μp_old(s_current), σp_old(s_current))}, the corrected physiological baseline parameters are calculated, where the corrected mean μp - update(s_current) is the standard deviation of the physiological characteristic. The formulas for calculating μp-update(s_current) and standard deviation σp-update(s_current) are: μp-update(s_current) = (1-λ)·μp_old(s_current) + λ·μp_new(s_current), σp-update(s_current) = (1-λ)·σp_old(s_current) + λ·σp_new(s_current); thus, the corrected physiological baseline is: Bp_update(s_current) = {(μp-update,σp-update(s_current))};

[0111] Using the same logic and formula as the above-mentioned independent correction of physiological baselines, the stationary subsequences of behavioral features B_stable and operational features O_stable are processed respectively to obtain the corrected behavioral baseline Bb_update(s_current) and operational baseline Bo_update(s_current).

[0112] The cross-dimensional correlation collaborative correction unit adjusts the baseline after independent correction in one dimension based on the changes in the correlation between multi-dimensional stationary data sequences; the specific analysis process is as follows:

[0113] The Pearson correlation coefficient method was used to calculate the physiological-behavioral correlation coefficient r_p-b, the physiological-operational correlation coefficient r_p-o, and the behavioral-operational correlation coefficient r_b-o, respectively. Historical cross-dimensional correlation coefficients r_old for the corresponding scenario s_current in the initial individual baseline set B0 were extracted, including r_p-b-old, r_p-o-old, and r_b-o-old. The corresponding change in correlation coefficient Δr = |r_new - r_old| was calculated, where r_new is the new correlation coefficient calculated based on stationary data. A preset correlation change threshold Δr0 was used. If Δr ≥ Δr0, collaborative correction was triggered; if Δr < Δr0, the correlation was determined to have no significant change, no collaborative correction was needed, and the baseline parameters after single-dimensional independent correction were retained.

[0114] In this embodiment, if Δr ≥ Δr0, then collaborative correction is triggered:

[0115] Taking the physiological-behavioral association as an example, if r_p-b-new > r_p-b-old (indicating an enhanced association between physiological and behavioral characteristics), then the adjustment range for the physiological baseline and the behavioral baseline is adjusted, as shown in the formula:

[0116] The physiological mean after co-correction is μp_co(s_current) = μp - update(scurrent)·(1+e·Δr);

[0117] The mean behavior after collaborative correction is μb_co(s_current) = μb - update(scurrent)·(1 + e·Δr); where e is the correlation influence coefficient (0 < e < 0.5, used to control the magnitude of collaborative correction).

[0118] If Δr < Δr0, the correlation is determined to be unchanged, no collaborative correction is required, and the baseline parameters after single-dimensional independent correction are retained.

[0119] Repeat the above process to complete the collaborative correction of the physiological-operational and behavioral-operational dimensions, and obtain the baselines of each dimension after collaborative correction (Bp_co(s_current), Bb_co(s_current), Bo_co(s_current).

[0120] The updated baseline set integration unit integrates the baselines of each dimension after independent correction in one dimension or collaborative correction across dimensions, forming the updated baseline group B_update(s_current) for the corresponding scenario, represented as:

[0121] B_update(s_current)={Bp_final(s_current),Bb_final(s_current),Bo_final(s_current)}, where Bp_final is the physiological baseline after independent or collaborative correction, Bb_final is the behavioral baseline after independent or collaborative correction, and Bo_final is the operational baseline after independent or collaborative correction; the updated baseline groups of all scenarios are integrated according to the structure of the initial individual baseline set B0 to form an updated individual baseline set Bnew covering all typical driving scenarios, and Bnew={B_update(s|s∈typical driving scenario set)}.

[0122] The driver status judgment module includes a scenario baseline call unit, a single-dimensional deviation value calculation unit, a multi-dimensional collaborative deviation value calculation unit, an abnormal status judgment unit, and a judgment result output unit;

[0123] The scenario baseline invocation unit receives the current driving scenario type s_current output by the driving scenario classification unit and the updated individual baseline set Bnew output by the individual baseline dynamic update module. It then calls the target scenario baseline group B_target(s_current) that matches s_current from the updated individual baseline set Bnew. The target scenario baseline group includes the updated physiological baseline Bp_final(s_current), behavioral baseline Bb_final(s_current), and operational baseline Bo_final(s_current).

[0124] The single-dimensional deviation calculation unit receives real-time multi-dimensional data output from the multi-dimensional real-time data acquisition unit and the target scene baseline group B_target(s_current) output from the scene baseline call unit, and calculates the single-dimensional deviation value of each feature according to the dimension. The specific analysis process is as follows:

[0125] With a preset deviation correction coefficient k1, and referring to the calculation formula of the real-time deviation in the multi-dimensional deviation calculation unit, calculate the single-dimensional deviation value Dp_single(t) of the physiological dimension, the single-dimensional deviation value Db_single(t) of the behavioral dimension, and the single-dimensional deviation value Do_single(t) of the operational dimension.

[0126] The multi-dimensional collaborative deviation calculation unit calculates the multi-dimensional collaborative deviation value based on the single-dimensional deviation value and the scenario-based weights of each dimension. It also incorporates the correction results of the inter-dimensional collaborative coefficients. The specific analysis is as follows:

[0127] A pre-defined scenario-based dimension weight library is used. For the current scenario s_current, the corresponding weights w_p(s_current), w_b(s_current), and w_o(s_current) are called, satisfying w_p(s_current) + w_b(s_current) + w_o(s_current) = 1, and the weight allocation matches the characteristics of the scenario. The initial multi-dimensional collaborative deviation value Dc_init(t) is calculated using the formula: Dc_init(t) = w_p(s_current)·Dp_single(t) + w_b(s_current)·Db_single(t) + w_o(s_current) ent)·Do_single(t); Based on the preset co-analysis time window T0, the synchronicity of the real-time single-dimensional deviation value is calculated using the Pearson correlation coefficient, thereby obtaining the physiological-operational co-coefficient Cp-o(t), the physiological-behavioral co-coefficient Cp-b(t), and the behavioral-operational co-coefficient Cb-o(t). The average of the three is taken as the final co-coefficient C(t); If C(t)≥0.6, then the coefficient of Dc_init(t) is corrected, and the corresponding formula is: Dc_final(t)=Dc_init(t)·(1+f·C(t)), where f represents the coefficient; If C(t)<0.6, then Dc_final(t)=Dc_init(t);

[0128] The abnormal state determination unit determines the driver's current state based on the single-dimensional deviation value and the corrected multi-dimensional collaborative deviation value, combined with a preset threshold. If Dp_single(t) < D1, Db_single(t) < D1, Do_single(t) < D1 and Dc_final(t) < D2, it is determined to be a "normal state", where D1 is the single-dimensional abnormal threshold and D2 is the multi-dimensional abnormal threshold.

[0129] If any single-dimensional deviation value is greater than or equal to D1 and Dc_final(t) < D2, it is judged as a "mildly abnormal state";

[0130] If Dc_final(t)≥D2 and C(t)<D3, it is determined to be a “moderately abnormal state”, where D3 is the synergistic enhancement threshold;

[0131] If Dc_final(t)≥D2 and C(t)≥D3, it is determined to be a "severe abnormal state";

[0132] The judgment result output unit receives the status judgment result output by the abnormal status judgment unit and outputs the judgment result in a preset format, which includes "current status type (normal / mild abnormal / moderate abnormal / severe abnormal), abnormal dimension identifier (such as "physiological dimension - heart rate abnormality"), single dimension deviation value details, multi-dimensional collaborative deviation value and collaborative coefficient".

[0133] A driver status multi-dimensional recognition device, including a multi-dimensional sensor group and a system early warning module;

[0134] The multi-dimensional sensor group includes physiological feature sensors, behavioral feature sensors, operational feature sensors, and scene recognition sensors, such as DMS cameras; the system warning module receives the judgment results output by the driver status judgment module and triggers corresponding warning interventions according to different abnormal status types. The warning interventions include light prompts, seat vibration, and voice warnings.

[0135] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0136] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-dimensional driver status recognition system, characterized in that: The system includes: an initial individual baseline construction module, a real-time data acquisition module, a smooth driving period identification module, an individual baseline dynamic update module, and a driver status judgment module; The initial individual baseline construction module, based on a preset baseline calibration process, collects multi-dimensional data of drivers when they first use the system in various typical driving scenarios, and generates an initial individual baseline set for the driver in the corresponding scenario based on the collected multi-dimensional data. The real-time data acquisition module collects real-time multi-dimensional data and real-time scene feature data of the driver during subsequent driving. The smooth driving period identification module performs hierarchical analysis on real-time multi-dimensional data and real-time scene feature data. First, it matches the corresponding scene baseline, and then combines the scene baseline to filter smooth driving segments in real-time multi-dimensional data, merging continuous smooth driving segments into smooth driving periods. The individual baseline dynamic update module periodically adjusts the corresponding baselines in the initial individual baseline set based on real-time multi-dimensional data during the stable driving period, forming an updated individual baseline set; the periodic adjustment includes independent correction of the feature parameters of each dimension baseline and collaborative correction of cross-dimensional correlation parameters. The individual baseline dynamic update module includes a single-dimensional baseline independent correction unit, a cross-dimensional correlation collaborative correction unit, and an updated baseline set integration unit. The single-dimensional baseline independent correction unit, based on the multi-dimensional stationary data sequence output by the stationary data extraction unit, independently corrects the baselines of the corresponding scenarios in the initial individual baseline set B_0 according to the physiological, behavioral, and operational dimensions. The cross-dimensional correlation collaborative correction unit adjusts the baseline after independent correction in one dimension based on the changes in the correlation relationship of multi-dimensional stationary data sequences; the specific analysis process is as follows: The Pearson correlation coefficient method was used to calculate the physiological-behavioral correlation coefficient r_p-b, the physiological-operational correlation coefficient r_p-o, and the behavioral-operational correlation coefficient r_b-o, respectively. The historical cross-dimensional correlation coefficient r_old for the corresponding scenario s_current in the initial individual baseline set B0 was extracted, including r_p-b-old, r_p-o-old, and r_b-o-old. The corresponding change in correlation coefficient Δr = |r_new - r_old| was calculated, where r_new is the new correlation coefficient calculated based on stationary data. A preset correlation change threshold Δr0 was used. If Δr ≥ Δr0, collaborative correction was triggered; if Δr < Δr0, no collaborative correction was required, and the baseline parameters after single-dimensional independent correction were retained. The updated baseline set integration unit integrates the baselines of each dimension after single-dimensional independent correction or cross-dimensional collaborative correction to form an updated baseline group for the corresponding scenario. The driver status judgment module compares the real-time multi-dimensional data with the updated individual baseline and the current scenario baseline, calculates the single-dimensional deviation value and multi-dimensional collaborative deviation value of each dimension data, and determines whether the driver's current status is abnormal based on the combined result of the two.

2. The driver status multi-dimensional recognition system according to claim 1, characterized in that: The initial individual baseline construction module includes a scene data acquisition unit, a baseline calculation unit, and a baseline set integration unit; The scenario data acquisition unit determines a set of typical driving scenarios, guides the driver to complete smooth driving operations in each scenario, and synchronously collects multi-dimensional data sequences during the process through multi-dimensional sensors. The typical driving scenario set includes urban road scenarios, highway scenarios, nighttime road scenarios, and rain and snow weather scenarios. The multi-dimensional data sequence includes physiological feature data sequence, behavioral feature data sequence, and operational feature data sequence. Among them, the physiological feature data sequence includes continuously collected heart rate data and blink duration data; the behavioral feature data sequence includes continuously collected eye gaze angle data and head rotation angle data; and the operational feature data sequence includes continuously collected steering wheel rotation angle data and accelerator pedal depth data. The baseline calculation unit performs statistical analysis on multi-dimensional data sequences under typical driving scenarios to obtain the mean and standard deviation of each feature parameter, generating scenario-specific physiological baselines, behavioral baselines, and operational baselines. The physiological baseline generated from the physiological feature data sequence is expressed as: Bp(s) = {(μp1(s),σp1(s)),(μp2(s),σp2(s)),...,(μpn(s),σpn(s))}, where s represents a typical driving scenario, μpi(s) is the mean of the i-th physiological feature in scenario s, and σpi(s) is the standard deviation of the i-th physiological feature in scenario s, with i ranging from 1 to n. Similarly, the behavioral feature data sequence and operational feature data sequence are processed in the same way to generate scenario-specific behavioral baselines Bb(s) and operational baselines Bo(s), and the expressions for the behavioral and operational baselines are consistent with those for the physiological baseline. The baseline set integration unit integrates the physiological baseline, behavioral baseline and operational baseline of each scenario to form an initial individual baseline set B0 containing multi-dimensional baselines corresponding to each scenario, and B0={Bp(s),Bb(s),Bo(s)|s∈typical driving scenario set}.

3. The driver status multi-dimensional recognition system according to claim 1, characterized in that: The real-time data acquisition module includes a scene environment parameter acquisition unit, a driving scene classification unit, and a multi-dimensional real-time data acquisition unit. The scene environment parameter acquisition unit collects real-time scene feature data through a scene recognition sensor. The real-time scene feature data includes road type parameters, light intensity parameters, and precipitation intensity parameters. The driving scene classification unit classifies and matches the real-time scene feature data acquired by the scene environment parameter acquisition unit, and identifies the current driving scene type based on the matching results. After the driving scene classification unit identifies the current driving scene type, the multi-dimensional real-time data acquisition unit simultaneously collects the driver's real-time physiological feature data, real-time behavioral feature data, and real-time operational feature data in that scene.

4. The driver status multi-dimensional recognition system according to claim 3, characterized in that: The driving scenario classification unit classifies and matches the real-time environmental parameters acquired by the scenario environment parameter acquisition unit, and identifies the current driving scenario type based on the matching results. The specific process is as follows: A scenario classification rule library is pre-built, which includes the threshold range of environmental parameters corresponding to each typical driving scenario. The threshold range of road type parameters includes the road width range, lane number range and speed limit range corresponding to urban roads. The threshold range of light intensity parameters includes the light intensity range corresponding to night roads. The threshold range of precipitation intensity parameters includes the precipitation intensity range corresponding to rain and snow weather. The system receives real-time scene feature data output from the scene environment parameter acquisition unit. This real-time scene feature data includes real-time road type parameter R, real-time illumination intensity parameter L, and real-time precipitation intensity parameter W. The system then matches the real-time road type parameter R with the threshold range of road type parameters for each scene in the scene classification rule base to obtain a road type matching result; matches the real-time illumination intensity parameter L with the threshold range of illumination intensity parameters for each scene in the scene classification rule base to obtain an illumination intensity matching result; and matches the real-time precipitation intensity parameter W with the threshold range of precipitation intensity parameters for each scene in the scene classification rule base to obtain a precipitation intensity matching result. Based on the above three matching results, the current driving scenario type is determined using a scenario matching degree calculation model. When all three matching results corresponding to a typical driving scenario meet the threshold requirements, or meet two core matching results and the third matching result is not conflicting, the current driving scenario type is determined to be that typical driving scenario. The scenario matching degree S(s) is calculated as follows: S(s) = α·S_R(s) + β·S_L(s) + γ·S_W(s), where s represents the typical driving scenario, S_R(s) is the matching degree between the real-time road type parameter R and the threshold range of the road type parameter in scenario s, and S_L(s) is the real-time road type parameter R. S_W(s) represents the matching degree between the light intensity parameter L and the threshold range of the light intensity parameter in scene s; S_W(s) represents the matching degree between the real-time precipitation intensity parameter W and the threshold range of the precipitation intensity parameter in scene s; α, β, and γ are the weighting coefficients of the road type matching degree, light intensity matching degree, and precipitation intensity matching degree, respectively, and α+β+γ=1; when S(s)≥0.8, the current driving scene type is determined to be scene s; when there are multiple scenes s with S(s)≥0.8, the scene with the largest S(s) is selected and the road type parameter is further refined based on S_R(s) corresponding to α, so that the scene with the largest S_R(s) is selected as the current driving scene type.

5. The driver status multi-dimensional recognition system according to claim 3, characterized in that: The smooth driving period identification module includes a scene baseline matching unit, a multi-dimensional deviation calculation unit, a smooth driving segment filtering unit, and a smooth driving period merging unit. The scenario baseline matching unit receives the current driving scenario type s_current output by the driving scenario classification unit, and calls the sub-scenario baseline group B(s_current) corresponding to the current driving scenario type from the initial individual baseline set B0 generated by the baseline set integration unit in the initial individual baseline construction module; wherein, the sub-scenario baseline group B(s_current) includes the physiological baseline Bp(s_current), behavioral baseline Bb(s_current), and operational baseline Bo(s_current) of the sub-scenario. The multi-dimensional deviation calculation unit receives real-time multi-dimensional data output by the multi-dimensional real-time data acquisition unit, and calculates the real-time deviation of each dimension in combination with the sub-scene baseline group B(s_current) called by the scene baseline matching unit. The smooth driving segment filtering unit filters smooth driving segments from real-time multi-dimensional data based on the real-time deviation of each dimension output by the multi-dimensional deviation calculation unit and the driving state fluctuation judgment conditions. The smooth driving period merging unit merges multiple smooth driving segments output by the smooth driving segment filtering unit to form a complete smooth driving period.

6. The driver status multi-dimensional recognition system according to claim 5, characterized in that: The specific process for calculating the real-time deviation of each dimension in the multi-dimensional deviation calculation unit is as follows: The physiological deviation Dp(t) between the real-time physiological characteristic number P(t) and the physiological baseline Bp(s_current) is calculated using the following formula: Dp(t) = (1 / n)∑ i∈[1,n] [|pi(t)-μ_pi(s_current)| / (k·σ_pi(s_current))], where pi(t) is the real-time value of the i-th physiological feature at time t, μ_pi(s_current) is the baseline mean of the i-th physiological feature in the s_current scenario, σ_pi(s_current) is the baseline standard deviation of the i-th physiological feature in the s_current scenario, and k is a preset deviation correction coefficient; using the same calculation logic and formula as the above physiological deviation, the behavioral deviation Db(t) between the real-time behavioral feature data B(t) and the behavioral baseline Bb(s_current), and the operational deviation Do(t) between the real-time operational feature data O(t) and the operational baseline Bo(s_current) are calculated respectively; The specific process of filtering smooth driving segments from real-time multi-dimensional data in the smooth driving segment filtering unit is as follows: The preset stable deviation threshold D and driving state fluctuation threshold Z are used. For each time point t, if Dp(t)≤D, Db(t)≤D and Do(t)≤D, and the variance of the real-time operation feature data O(t) in three consecutive time points Var(O(t),O(t-1),O(t-2))≤Z, then time t is determined to be a stable driving point. The continuous data segments are formed by combining consecutive timestamps of smooth driving points. When the duration of a continuous data segment is greater than or equal to a preset segment duration threshold, the continuous data segment is marked as a smooth driving segment. The specific analysis process of the smooth driving period merging unit is as follows: The preset segment interval threshold ε and the shortest duration threshold θ are used. If the time interval between two adjacent smooth driving segments is less than or equal to the segment interval threshold ε, the two segments are merged into a new smooth driving segment. The above merging operation is repeated until the interval between all adjacent segments is greater than the segment interval threshold ε. Segments with a duration greater than or equal to the shortest duration threshold θ after merging are selected and marked as the final smooth driving period.

7. The driver status multi-dimensional recognition system according to claim 1, characterized in that: The individual baseline dynamic update module also includes a stable data extraction unit; The stable data extraction unit receives the stable driving period output by the stable driving period identification module and the current driving scenario type s_current output by the driving scenario classification unit, and extracts the multi-dimensional stable data sequence of the current driving scenario type within the stable driving period from the real-time multi-dimensional data. The multi-dimensional stable data sequence includes the physiological feature stable subsequence P_stable, the behavioral feature stable subsequence B_stable, and the operational feature stable subsequence O_stable. The single-dimensional baseline independent correction unit, based on the multi-dimensional stationary data sequence output by the stationary data extraction unit, independently corrects the baselines of the corresponding scenarios in the initial individual baseline set B_0 according to physiological, behavioral, and operational dimensions; the specific analysis process is as follows: A baseline update period T and a correction coefficient λ are preset, where 0 < λ < 1. When the system runtime reaches the update period T, an independent correction process is triggered to perform statistical analysis on the stationary subsequence P_stable of the physiological characteristics, calculating the stationary mean μp_new(s_current) and stationary standard deviation σp_new(s_current) of the i-th physiological characteristic. Based on the initial physiological baseline Bp_old(s_current) = {(μp_old(s_current), σp_old(s_current))}, the corrected physiological baseline parameters are calculated, where the corrected mean μp - update(s_current) is the standard deviation of the physiological characteristic. The formulas for calculating μp-update(s_current) and standard deviation σp-update(s_current) are: μp-update(s_current) = (1-λ)·μp_old(s_current) + λ·μp_new(s_current), σp-update(s_current) = (1-λ)·σp_old(s_current) + λ·σp_new(s_current); thus, the corrected physiological baseline is: Bp_update(s_current) = {(μp-update,σp-update(s_current))}; Using the same logic and formula as the above-mentioned independent correction of physiological baselines, the stationary subsequences of behavioral features B_stable and operational features O_stable are processed respectively to obtain the corrected behavioral baseline Bb_update(s_current) and operational baseline Bo_update(s_current). The updated baseline set integration unit integrates the baselines of each dimension after independent correction in one dimension or collaborative correction across dimensions, forming the updated baseline group B_update(s_current) for the corresponding scenario, represented as: B_update(s_current)={Bp_final(s_current),Bb_final(s_current),Bo_final(s_current)}, where Bp_final is the physiological baseline after independent or collaborative correction, Bb_final is the behavioral baseline after independent or collaborative correction, and Bo_final is the operational baseline after independent or collaborative correction; the updated baseline groups of all scenarios are integrated according to the structure of the initial individual baseline set B0 to form an updated individual baseline set Bnew covering all typical driving scenarios, and Bnew={B_update(s|s∈typical driving scenario set)}.

8. A driver status multi-dimensional recognition system according to claim 7, characterized in that: The driver status judgment module includes a scenario baseline call unit, a single-dimensional deviation value calculation unit, a multi-dimensional collaborative deviation value calculation unit, an abnormal status judgment unit, and a judgment result output unit. The scenario baseline invocation unit receives the current driving scenario type s_current output by the driving scenario classification unit and the updated individual baseline set Bnew output by the individual baseline dynamic update module. It then calls the target scenario baseline group B_target(s_current) that matches s_current from the updated individual baseline set Bnew. The target scenario baseline group includes the updated physiological baseline Bp_final(s_current), behavioral baseline Bb_final(s_current), and operational baseline Bo_final(s_current). The single-dimensional deviation calculation unit receives real-time multi-dimensional data output from the multi-dimensional real-time data acquisition unit and the target scene baseline group B_target(s_current) output from the scene baseline call unit, and calculates the single-dimensional deviation value of each feature according to the dimension. The specific analysis process is as follows: With a preset deviation correction coefficient k1, and referring to the calculation formula of the real-time deviation in the multi-dimensional deviation calculation unit, calculate the single-dimensional deviation value Dp_single(t) of the physiological dimension, the single-dimensional deviation value Db_single(t) of the behavioral dimension, and the single-dimensional deviation value Do_single(t) of the operational dimension. The multi-dimensional collaborative deviation calculation unit calculates the multi-dimensional collaborative deviation value based on the single-dimensional deviation value and the scenario-based weights of each dimension. It also incorporates the correction results of the inter-dimensional collaborative coefficients. The specific analysis is as follows: A pre-defined scenario-based dimension weight library is used. For the current scenario s_current, the corresponding weights w_p(s_current), w_b(s_current), and w_o(s_current) are called, satisfying w_p(s_current) + w_b(s_current) + w_o(s_current) = 1, and the weight allocation matches the characteristics of the scenario. The initial multi-dimensional collaborative deviation value Dc_init(t) is calculated using the formula: Dc_init(t) = w_p(s_current)·Dp_single(t) + w_b(s_current)·Db_single(t) + w_o(s_current) ent)·Do_single(t); Based on the preset co-analysis time window T0, the synchronicity of the real-time single-dimensional deviation value is calculated using the Pearson correlation coefficient, thereby obtaining the physiological-operational co-coefficient Cp-o(t), the physiological-behavioral co-coefficient Cp-b(t), and the behavioral-operational co-coefficient Cb-o(t). The average of the three is taken as the final co-coefficient C(t); If C(t)≥0.6, then the coefficient of Dc_init(t) is corrected, and the corresponding formula is: Dc_final(t)=Dc_init(t)·(1+f·C(t)), where f represents the coefficient; If C(t)<0.6, then Dc_final(t)=Dc_init(t); The abnormal state determination unit determines the driver's current state based on the single-dimensional deviation value and the corrected multi-dimensional collaborative deviation value, combined with a preset threshold. If Dp_single(t) < D1, Db_single(t) < D1, Do_single(t) < D1 and Dc_final(t) < D2, it is determined to be a "normal state", where D1 is the single-dimensional abnormal threshold and D2 is the multi-dimensional abnormal threshold. If any single-dimensional deviation value is greater than or equal to D1 and Dc_final(t) < D2, it is judged as "mild abnormal state"; If Dc_final(t)≥D2 and C(t)<D3, it is determined to be a "moderately abnormal state", where D3 is the synergistic enhancement threshold; If Dc_final(t)≥D2 and C(t)≥D3, it is determined to be a "severe abnormal state"; The judgment result output unit receives the status judgment result output by the abnormal status judgment unit and outputs the judgment result containing "current status type, abnormal dimension identifier, single dimension deviation value details, multi-dimensional collaborative deviation value and collaborative coefficient" in a preset format.

9. A driver status multi-dimensional recognition device, characterized in that, The driver status multi-dimensional recognition system according to any one of claims 1-8 further includes a multi-dimensional sensor group and a system early warning module; The multi-dimensional sensor group includes physiological feature sensors, behavioral feature sensors, operational feature sensors, and scene recognition sensors; the system early warning module receives the judgment result output by the driver state judgment module and triggers corresponding early warning interventions according to different abnormal state types, including light prompts, seat vibration, and voice warnings.

Images