A driver alert method, system, and vehicle based on seat massage

By using multimodal data scoring and responsiveness feedback adjustment, the seat massage parameters are adaptively adjusted, solving the problem that existing massage solutions cannot adapt to individual differences and fatigue changes, thus achieving personalized and continuous effectiveness of driver fatigue alerts.

CN123104497APending Publication Date: 2026-11-17CHERY AUTOMOBILE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610973972.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-01
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing technologies, driver fatigue reminder methods based on seat massage cannot adaptively adjust massage parameters according to individual differences and changes in driver fatigue levels, resulting in poor effects for some people and overstimulation for others. Furthermore, the effectiveness of the reminder cannot be judged, leading to the problem of blind execution.

Method used

By acquiring multimodal fatigue characteristic data, performing weighted scoring calculations, determining the fatigue level, and setting massage parameters according to the level; after the massage, the responsiveness is reassessed, and the massage intensity and frequency are adaptively adjusted, linking multiple systems for collaborative reminders.

Benefits of technology

It enables personalized adaptation of massage parameters, improving the effectiveness and long-term effects of fatigue reminders, and avoiding discomfort and diminishing effects caused by fixed parameters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123104497A_ABST
    Figure CN123104497A_ABST
Patent Text Reader

Abstract

This invention provides a driver alert method, system, and vehicle based on seat massage, relating to the field of vehicle safety technology. The method acquires multimodal fatigue characteristic data, performs weighted scoring to obtain a comprehensive fatigue score and determine the fatigue level. After executing a massage alert corresponding to the level, the responsiveness is calculated to quantify the alert effect. When the responsiveness is below the effective alert threshold, the massage intensity and / or frequency are adaptively adjusted, forming a closed-loop control of detection, execution, evaluation, and optimization. This invention achieves automatic adaptation of massage parameters to individual driver differences and state changes through a responsiveness feedback mechanism, effectively solving the problem of inconsistent alert effects and attenuation over time with fixed parameter schemes, significantly improving the personalized adaptability and long-term effectiveness of fatigue driving alerts.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of vehicle safety technology, and in particular to a driver reminder method, system, and vehicle based on seat massage. Background Technology

[0002] Fatigue driving is one of the main causes of traffic accidents, especially during long-distance driving. Current technology uses cameras to monitor driver facial features to detect fatigue and triggers seat massage for tactile alerts, which is more effective than traditional warnings. However, existing solutions use fixed-parameter massage, meaning that regardless of individual driver differences, changes in fatigue levels, or adaptation after repeated reminders, the same massage intensity and waveform are applied. This one-size-fits-all approach has significant drawbacks: different drivers have sensitivities to tactile stimulation that can vary by several times; fixed parameters are insufficient for some and overstimulating for others; even the same driver may develop sensory adaptation to the same stimulus after prolonged driving, leading to a continuous decline in the effectiveness of the reminder; and the system cannot determine whether the reminder is truly effective, remaining in a state of blind execution. Summary of the Invention

[0003] To address the technical problems existing in the background art, the present invention provides a driver reminder method, system, and vehicle based on seat massage.

[0004] To achieve the above objectives, a first aspect of the present invention provides a driver reminder method based on seat massage, comprising: Obtain fatigue characteristic data for at least two different modes; The fatigue characteristic data are weighted and scored to obtain a comprehensive fatigue score. ; Based on the fatigue comprehensive score The fatigue level is determined by comparing it with a preset threshold. Based on the fatigue level, determine the corresponding seat massage parameters and control the seat massage module to perform massage reminders corresponding to the fatigue level; After the massage reminder is activated, the driver's fatigue characteristic data is reacquired, and the responsiveness is calculated. The responsiveness The fatigue comprehensive score is calculated based on the fatigue comprehensive score at the moment of massage triggering and the fatigue comprehensive score within the preset response evaluation time window after massage triggering; When the responsiveness When the intensity of the massage falls below the preset effective reminder threshold, the massage intensity and / or frequency will be adjusted adaptively.

[0005] Furthermore, the fatigue feature data is weighted and scored to obtain a comprehensive fatigue score. ,include: Obtain visual fatigue characteristic data, steering wheel behavior characteristic data, short-term fatigue accumulation characteristic data, and continuous driving duration data, and calculate visual fatigue characteristic scores for each. Steering wheel behavior characteristic score Short-term fatigue cumulative score and continuous driving time correction factor ; The fatigue comprehensive score S is calculated according to the weighted scoring formula: ; in, For real-time sampling time, This refers to the duration of continuous driving during that session. , , , The weighting coefficients are satisfied. , , , , The value range is [0,1].

[0006] Furthermore, based on the comparison between the comprehensive fatigue score S and a preset threshold, the fatigue level is determined, including: When the fatigue comprehensive score When the patient is in the first rating range, it is determined to be level one fatigue. The massage waveform is set to vibration mode and the massage intensity is set to low intensity. When the fatigue comprehensive score When the patient is in the second rating range, it is determined to be the second level of fatigue. The massage waveform is set to the pressing mode and the massage intensity is set to the medium intensity. When the fatigue comprehensive score When the body is in the third rating range, it is judged as level three fatigue. The massage waveform is set to tapping mode, the massage intensity is set to high intensity, and sound prompts or voice warnings are played simultaneously. Among them, the fatigue comprehensive score thresholds corresponding to the first, second, and third score intervals increase sequentially, with the third level of fatigue corresponding to the highest degree of fatigue.

[0007] Furthermore, when the responsiveness The response rate is less than 0 or after multiple consecutive massage reminders. If the threshold is still below the effective alert threshold, a multi-system coordinated alert is triggered, including: The system can be linked to the air conditioning system to lower the interior temperature or turn on the blower function. The linked audio system plays a high-volume alert tone or voice announcement; The integrated navigation system pushes location information and navigation routes for nearby service areas, parking lots, or rest areas onto the vehicle's infotainment screen.

[0008] Furthermore, the responsiveness The calculation formula is: ,in, The overall fatigue score is based on the moment the massage is triggered. The response evaluation time window after massage triggering. The overall fatigue score.

[0009] Furthermore, the adaptive adjustment of the massage intensity This includes: calculating the adjusted massage intensity based on the intensity adaptive adjustment formula. : ; in, For the first Intensity value triggered by the massage For the first The intensity value adjusted upon triggering the massage. For the first The measured response rate of a single massage. For target responsiveness, The step size coefficient is used to adjust the intensity. For the amplitude limiting function, For minimum safety strength, For maximum safety strength.

[0010] Furthermore, the adaptive adjustment of the massage frequency ,include: When continuous The responsiveness after secondary intensity adjustment The effective reminder threshold has not yet been reached. At that time, the adjusted massage frequency is calculated according to the frequency adaptive adjustment formula. : ; in, This is the frequency adjustment enable coefficient. For frequency adjustment step size, For symbolic functions, For the minimum frequency, This is the maximum frequency.

[0011] Furthermore, the triggering of the massage waveform switching or massage area rotation includes: Calculate the habituation inhibition factor: ; in, This refers to the number of consecutive uses of the current massage parameter combination. This is the habituation attenuation coefficient; based on the habituation inhibition factor... Calculate the actual execution intensity: ; in, The current intensity value. For actual enforcement intensity, This is the habituation compensation coefficient; when the habituation inhibition factor... Below the preset habituation threshold At the same time, the massage waveforms are switched in the order of sine wave, trapezoidal wave, and pulse wave, or the massage areas are rotated between the waist, back, and shoulders.

[0012] Furthermore, it also includes: if the number of consecutive uses of the current massage parameter combination reaches a preset threshold, triggering a switch of massage waveform or rotation of massage areas.

[0013] A second aspect of the present invention provides a driver alert system based on seat massage, comprising: The fatigue detection module includes an in-vehicle camera and a steering wheel angle sensor, used to acquire fatigue characteristic data of at least two different modes; The scoring calculation module is used to perform weighted scoring calculations on the multimodal fatigue feature data acquired by the fatigue detection module to obtain a comprehensive fatigue score. ; The grading and determination module is used to determine the fatigue comprehensive score. The fatigue level is output by comparing it with a preset threshold. The massage control module is used to determine the massage parameter vector based on the fatigue level. Send control commands to the seat massage actuator; The responsiveness assessment module is used to preset a time window after the massage reminder is activated. Reacquire driver fatigue characteristic data and calculate response. ; The parameter adjustment module includes an intensity adjustment unit, a frequency adjustment unit, and a habituation suppression unit. The intensity adjustment unit is used to adjust the response based on the response. Adjust massage intensity The frequency adjustment unit is used to adjust the frequency according to the response. Adjust massage frequency The habituation inhibition unit is used to calculate the habituation inhibition factor. And compensate for the intensity of the massage.

[0014] Furthermore, it also includes a linkage control module, which communicates with the air conditioning system, audio system, and navigation system via vehicle Ethernet or CAN bus. When the responsiveness is lower than the effective reminder threshold, the linkage control module triggers multi-system collaborative intervention.

[0015] Furthermore, the in-vehicle camera of the fatigue detection module is an infrared camera, used to acquire the driver's facial expressions, eye state, and head posture; the steering wheel angle sensor is used to acquire the driver's grip force changes and steering operation data; the fatigue detection module also includes a timing unit for recording continuous driving time.

[0016] Furthermore, the seat massage actuator includes a lumbar massage area located at the lower part of the seat back, a back massage area located in the middle of the seat back, and a shoulder massage area located at the upper part of the seat back. The three massage areas can be controlled independently or activated in combination. The massage control module controls the three massage areas to achieve the rotation of massage areas.

[0017] A third aspect of the invention provides a vehicle including the driver alert system based on seat massage provided in the second aspect of the invention.

[0018] Compared with existing technologies, the present invention provides a driver reminder method, system, and vehicle based on seat massage. By re-acquiring fatigue characteristic data and calculating the responsiveness after each massage reminder, the actual effect of the reminder is quantified. Then, the intensity and frequency of subsequent massages are dynamically adjusted according to the deviation between the responsiveness and the target value, forming a complete closed-loop control of detection, execution, evaluation, and optimization. This achieves adaptive convergence of massage parameters with individual differences and state changes of the driver, effectively overcoming the shortcomings of fixed parameter schemes that have different effects on different drivers and whose effects on the same driver decay over time. This significantly improves the personalized adaptability and long-term effectiveness of fatigue reminders. Attached Figure Description

[0019] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.

[0020] Figure 1 This is a flowchart of a driver reminder method based on seat massage provided in Embodiment 1 of the present invention.

[0021] Figure 2 This is a flowchart of the multimodal fatigue scoring calculation provided in Embodiment 1 of the present invention.

[0022] Figure 3 This is a flowchart of the adaptive adjustment of massage parameters provided in Embodiment 1 of the present invention.

[0023] Figure 4 This is a flowchart of habituation suppression and waveform switching provided in Embodiment 1 of the present invention.

[0024] Figure 5 This is an architecture diagram of a driver reminder system based on seat massage provided in Embodiment 2 of the present invention.

[0025] Figure 6 This is a diagram showing the internal structure of the parameter adjustment module provided in Embodiment 2 of the present invention. Detailed Implementation

[0026] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the invention. In the description of the embodiments of the present invention, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "this embodiment" should be understood as "at least one embodiment". 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 indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this embodiment, unless otherwise stated, "a plurality of" means two or more.

[0027] Before describing specific embodiments of the present invention, the key terms involved in the present invention will first be explained: 1. A Driver Monitoring System (DMS) is an in-vehicle active safety system that monitors the driver's condition in real time using in-vehicle cameras or sensors. This system can detect physiological characteristics such as the driver's facial orientation, eye opening and closing, and head posture, and can identify and issue warnings for abnormal states such as driver inattention and fatigue.

[0028] 2. PERCLOS (Percentage of Eye Closure) is a standardized indicator for measuring driver eye fatigue. It is defined as the percentage of time the eyelids are closed for more than 80% of the time within a specific time window. Proposed by the Federal Highway Administration (FHWA), this indicator is widely used in the field of driver fatigue detection and is currently one of the most representative visual fatigue indicators recognized by both academia and industry.

[0029] 3. Multimodal sensing refers to the technical method of simultaneously acquiring and analyzing the state of the same target using two or more different types of sensors or signal sources. In the field of driver state monitoring, multimodal sensing typically integrates multiple information such as visual signals (camera images), behavioral signals (steering wheel sensor data), physiological signals (heart rate, skin conductance, etc.), and environmental signals (vehicle speed, time, etc.) to improve the accuracy and robustness of state recognition.

[0030] 4. Sensory adaptation is a fundamental concept in neuroscience, referring to the physiological phenomenon where the sensitivity of a sensory organ to a stimulus gradually decreases over time when it is continuously subjected to the same stimulus. Sensory adaptation exists in multiple sensory channels, including vision, hearing, and touch, and is a universal adaptive mechanism of the human nervous system to continuous stimulation. In the field of touch, sensory adaptation is also known as habituation, and it is an important factor affecting the effectiveness of repetitive tactile cues.

[0031] 5. A clip function is a commonly used nonlinear function in signal processing and control systems. It is defined as follows: when the input value is below the lower limit, the output equals the lower limit; when the input value is above the upper limit, the output equals the upper limit; and when the input value is between the upper and lower limits, the output equals the input value. Clip functions are used to constrain signals or control quantities within a safe operating range, preventing system anomalies or hardware damage caused by parameters exceeding limits.

[0032] Example 1 like Figure 1 This embodiment provides a driver reminder method based on seat massage, including: S1. Obtain fatigue characteristic data for at least two different modes.

[0033] In this step, a multimodal fatigue detection mechanism is employed, integrating the following signal sources for fatigue assessment: Real-time facial images of the driver are acquired via an in-vehicle camera (preferably an infrared camera), extracting visual fatigue features such as facial expressions, eye states (e.g., eyelid opening and closing, blinking frequency), and head posture (e.g., nodding); real-time steering operation data of the driver is acquired via a steering wheel angle sensor, including steering wheel angle changes (e.g., steering wheel angle standard deviation, used to detect abnormal operations such as swerving and unstable steering) and grip force changes; simultaneously, the driver's continuous driving time is recorded via a timing unit. These three different modalities (visual modality, operational behavior modality, and temporal modality) of fatigue feature data collectively constitute multi-source detection data, providing a multi-dimensional information foundation for subsequent comprehensive fatigue assessment. This method does not rely on a single fatigue behavior trigger but rather accumulates scores for multiple features. Each feature is assigned a different weight based on its fatigue indication intensity, and an alert is triggered only when the accumulated score exceeds a preset threshold, effectively reducing the false alarm rate.

[0034] S2. Perform weighted scoring on the fatigue characteristic data to obtain a comprehensive fatigue score. .

[0035] like Figure 2 As shown, in this step, feature extraction and quantification scoring are first performed on the modal data obtained in step S1. Specifically: (1) Visual fatigue characteristic score Based on driver facial image recognition, the duration of eyelid closure, yawning, and nodding actions are analyzed. The proportion of eyelid closure duration (PERCLOS value), the frequency of yawning, and the number of nodding actions per unit time are calculated. These three factors are then weighted and combined to obtain a visual fatigue feature score. The value range is [0,1]; (2) Steering wheel behavior characteristics score Based on steering wheel angle sensor data, the standard deviation of steering wheel angle (characterizing the stability of the driving trajectory and used to detect abnormal operations such as swerving and unstable steering) and the rate of change in grip force (characterizing the driver's focus on steering wheel operation) are calculated per unit time. After normalization, the two are weighted and synthesized to obtain the steering wheel behavior feature score. The value range is [0,1]; (3) Short-term fatigue cumulative score : A sliding time window based on a preset length (window length) (e.g., 5 minutes), the frequency of occurrence of visual fatigue characteristics and driving behavior characteristics within the time window is counted, and the frequency of occurrence is normalized to obtain a short-term fatigue cumulative score. The value range is [0,1]; (4) Continuous driving duration correction factor Based on continuous driving time (Unit: minutes, and variables representing real-time sampling times) (The two have different meanings and are independent of each other) Calculate the duration correction factor ,in, Set a maximum safe driving time threshold (e.g., 120 minutes). The value range is [0,1].

[0036] Then, calculate the overall fatigue score according to the weighted scoring formula. : ; in, , , , The weight coefficients for each feature satisfy the following conditions: ,and That is, visual features have the highest weight; A smaller value is usually chosen (recommended range 0.10 to 0.15) to avoid the single factor of continuous driving time dominating the scoring results. The weighted sum of each component shows that each component's value is within the range [0,1], and the sum of the weights is 1. Therefore... The value range is [0,1], which is a dimensionless normalized score.

[0037] The above weighting coefficients , , , The value is determined in the following way: sample data of different drivers under various fatigue states are obtained, and the fatigue subjective evaluation scale (such as the Karolinska Sleepiness Score KSS) is used as the standard. The contribution of each feature is obtained by fitting the data using methods such as multiple linear regression or neural networks. Based on this, the weight coefficients are determined and stored in the vehicle memory as the factory default value. At the same time, personalized adjustments are allowed through the vehicle interface.

[0038] By using the weighted fusion of the above multi-dimensional features, compared with the single feature threshold judgment method, the false alarm rate caused by a single occasional fatigue feature (such as an accidental yawn) can be significantly reduced, and a continuous, quantitative and accurate assessment of the driver's fatigue level can be achieved.

[0039] like Figure 2 As shown, when the overall fatigue score is lower than the preset minimum effective threshold, no alert is triggered, and monitoring continues; only when the overall fatigue score exceeds the fatigue threshold will the subsequent fatigue level determination process begin.

[0040] S3. Based on the aforementioned comprehensive fatigue score The fatigue level is determined by comparing it with a preset threshold.

[0041] In this step, the fatigue comprehensive score calculated in step S2 will be used. Compared with the three preset thresholds (mild threshold) Moderate threshold Severe threshold ,satisfy Compare them. When If the fatigue level is determined to be not high or only slightly high, no alert will be triggered; when... When it is determined to be level one fatigue; when It was determined to be level two fatigue at that time; It was determined to be level three fatigue, which corresponds to the highest level of fatigue.

[0042] The above , , The values ​​of the three thresholds are determined as follows: First, the weighted scoring formula in step S2 is used to score and calculate the driver fatigue data obtained from a large number of real vehicles, resulting in a comprehensive fatigue score. The distribution range was determined; then, using the Subjective Evaluation Scale for Fatigue (KSS, 8-9 points for severe drowsiness) as a benchmark, the Receiver Operating Characteristic (ROC) curve method was employed to select the scoring cutoff points that optimally maximized both sensitivity and specificity (maximum Youden index) for fatigue detection, which were then used as the mild thresholds. Moderate threshold and severe threshold This information is stored as a factory default value in the vehicle's memory; it also allows for personalized adjustments via the vehicle's infotainment interface. For example, it can be set... , , .

[0043] By establishing a three-level fatigue assessment system, it is possible to accurately distinguish different stages of driver fatigue, providing a basis for the precise implementation of subsequent differentiated reminder strategies.

[0044] S4. Based on the fatigue level, determine the corresponding seat massage parameters and control the seat massage module to execute massage reminders corresponding to the fatigue level.

[0045] In this step, the corresponding massage parameter vector is determined based on the fatigue level determined in step S3. . Specifically: When the driver is assessed as experiencing Level 1 fatigue, the massage waveform is set to vibration mode (sine wave), the massage intensity is set to a low level (e.g., normalized value 0.2-0.4), and the reminder duration is 30 seconds to 1 minute. Gentle vibration is used to avoid excessive interference with the driver. If the seat massage is not activated, it is activated; if the seat massage is activated but not in vibration mode or the intensity is higher than low, it is adjusted to vibration mode and the intensity is set to low.

[0046] When the fatigue level is determined to be Level 2, set the massage waveform to pressure mode (trapezoidal wave), the massage intensity to medium (e.g., normalized value 0.4-0.7), and the alert duration to 1 to 2 minutes, providing moderate tactile stimulation to enhance alertness. If the seat massage is not activated, activate it; if the seat massage is activated but not in pressure mode or the intensity is below medium, adjust it to pressure mode and set the intensity to medium.

[0047] When fatigue is detected as Level 3, the massage waveform is set to a tapping mode (pulse wave), and the massage intensity is set to a high level (e.g., normalized value 0.7-1.0). Simultaneously, an audio prompt or voice warning is played for at least 2 minutes. Specifically, this includes the following scenarios: If the seat massage is not activated, it is activated and adjusted to tapping mode and highest intensity, accompanied by an audio prompt or voice warning; if the seat massage is activated but not in tapping mode or at the highest intensity, it is adjusted to tapping mode and highest intensity, accompanied by an audio prompt or voice warning; if the seat massage is activated in tapping mode and at the highest intensity, it is deactivated for 5 seconds and then reactivated, maintaining the tapping mode and highest intensity, accompanied by an audio prompt or voice warning. Each time the seat massage is activated due to fatigue alert, it is not automatically deactivated until the user manually deactivates it. Each activation triggers a separate functional cycle.

[0048] Through the above-mentioned differentiated tiered strategy, precise and humane interventions were achieved, ranging from mild and gentle reminders to moderate and intense stimulation and strong arousal.

[0049] S5. After the massage reminder is executed, reacquire the driver's fatigue characteristic data and calculate the responsiveness. The responsiveness The fatigue score is calculated based on the fatigue score at the moment of massage triggering and the fatigue score within the preset response evaluation time window after massage triggering.

[0050] In this step, after the massage reminder is activated in step S4, a preset response evaluation time window is reached. (e.g., 30 seconds), repeat steps S1 and S2 to acquire fatigue characteristic data and calculate the comprehensive fatigue score, obtaining the results after the massage trigger. Overall fatigue score at all times Then, the responsiveness of this massage reminder is calculated according to the responsiveness formula. : ; in, This is a comprehensive fatigue score at the moment the massage is triggered, serving as the baseline value for determining the driver's fatigue level when a massage reminder needs to be triggered. Below the minimum effective threshold If the value is 0.05, it is determined that the current fatigue level is not sufficient to trigger an alert, and no response rate calculation is performed to avoid calculation anomalies caused by the denominator approaching zero.

[0051] The above minimum effective threshold The value of is determined by calculating a comprehensive fatigue score from a large number of drivers in non-fatigue states. The distribution mean and standard deviation will Set the value to the mean plus twice the standard deviation (i.e., the upper limit of 95% confidence) as the benchmark value to distinguish between normal awake state and detectable fatigue state; for example, it can be set to 0.05.

[0052] responsiveness This characterizes the relative extent to which massage reminders reduced the overall fatigue score: A higher value indicates a greater decrease in fatigue score and a better alert effect; when When the fatigue score remains unchanged, the alert is ineffective; when This indicates that the fatigue score actually increases, and the driver's fatigue level continues to worsen during the massage, which is a dangerous situation. Define an effective alert threshold. (e.g., 0.3, meaning a fatigue score reduction of at least 30%) and invalid alert threshold. (e.g., 0.1, meaning the fatigue score decreases by less than 10%): When The reminder is deemed valid at this time, and no parameter adjustment is required; when If the alert is deemed severely ineffective, immediate escalation of intervention is required; when If the current alert is deemed insufficient but not yet dangerous, adaptive adjustments to the massage parameters are needed to improve subsequent alert effectiveness. If the driver's fatigue level continues to worsen during the massage, it indicates that the current driving condition has entered a high-risk zone. It is necessary to immediately trigger a multi-system coordinated alert and advise the driver to stop and rest as soon as possible.

[0053] The above effective reminder threshold Invalid reminder threshold The value is determined by the following method: based on the fatigue level assessment standard, through conversion. This indicates that each reminder is expected to reduce the fatigue score by more than 30%, and this value corresponds to the score change required to decrease the fatigue level by one level. A fatigue score reduction of less than 10% is considered a warning to be essentially ineffective. This value is the minimum clinically significant difference (MCID) that distinguishes between "effective" and "ineffective." Analysis of the driver's subjective perception of fatigue improvement... The correspondence between the values ​​is determined; the above thresholds are stored as factory default values ​​in the vehicle memory, and can also be adjusted through the vehicle interface.

[0054] By introducing a responsiveness index, an objective quantitative assessment of the actual effect of each massage reminder is achieved, providing a data foundation for subsequent adaptive parameter adjustment and upgrading this method from blind execution to intelligent control with feedback.

[0055] S6, when the response degree When the intensity of the massage falls below the preset effective reminder threshold, the massage intensity and / or frequency will be adjusted adaptively.

[0056] like Figure 3 As shown, in this step, the response degree calculated in step S5 is used. With preset effective reminder threshold Comparison. When If the reminder is valid, the current massage parameters will remain unchanged; if... If the system determines that the reminder is insufficient, it will trigger an adaptive adjustment of the parameters.

[0057] (1) Intensity adaptive adjustment: Massage intensity is the most direct parameter affecting the reminder effect; therefore, the intensity is first adaptively adjusted. The intensity value for the next massage is calculated according to the adaptive adjustment formula: ; in, For the first Intensity value triggered by the massage For the first The intensity value adjusted upon triggering the massage. For the first The measured response rate of a single massage. The target improvement rate (i.e., the expected reduction in fatigue score for each reminder). The step size coefficient is used to adjust the intensity. For the amplitude limiting function, For minimum safety strength, For maximum safety strength.

[0058] The above target improvement rate The value of is determined in the following way: Set to higher The expected value, exemplarily This indicates the expectation that each reminder will reduce the fatigue score by at least 50%, ensuring that the driver can recover from fatigue to a relatively alert and safe driving state. The value of the intensity adjustment step size coefficient must meet the following requirements. To ensure that the adjustment range does not exceed the safety constraint of 30%, exemplarily . and The hardware safety limits of the massage actuator are determined by the physical characteristics of the actuator, exemplarily... (Below this value, the driving force of the massager is insufficient to guarantee effective stimulation). (Corresponding to the actuator's rated maximum driving force). All of the above parameters are stored as factory default values ​​in the vehicle's onboard memory.

[0059] The core logic of this formula is: when the measured response... Below the target improvement rate When the intensity is insufficient to achieve the desired alert effect, the intensity value should be increased; conversely, when the intensity is insufficient, the intensity value should be increased. Higher than If the intensity is too high, it indicates that the massage may cause overstimulation, and the intensity should be appropriately reduced. Through this negative feedback adjustment mechanism, the massage intensity can be gradually converged to the optimal value that is most suitable for the current driver.

[0060] (2) Frequency adaptive adjustment: If the responsiveness still fails to meet the standard after multiple intensity adjustments, it indicates that simply adjusting the intensity is insufficient and further adjustment of the massage frequency is necessary. A tiered adjustment strategy, adjusting the intensity first and then the frequency, is adopted to avoid instability caused by adjusting multiple parameters simultaneously.

[0061] like Figure 3 As shown, when continuous Response after two (e.g., 2) intensity adjustments The effective reminder threshold has not yet been reached. When the frequency adjustment is initiated, the adjusted frequency value is calculated according to the frequency adaptive adjustment formula: ; in, and The first Second and third Frequency value when a massage is triggered. Frequency adjustment enable coefficient (when continuous) Even after intensity adjustment, the response still did not reach the target. hour, Initiate frequency adjustment; otherwise (without adjusting the frequency) For symbolic functions, For frequency adjustment step size, and These are the minimum and maximum frequencies, respectively. (Sign function) This determines the direction of frequency adjustment: when hour, Increase the frequency to enhance the intensity of the stimulation; when hour, Reduce the frequency appropriately to avoid overstimulation.

[0062] The above , , , The value of is determined in the following way: These are empirical values, aiming to strike a balance between rapid response and system stability. The frequency per minute strikes a balance between adjusting sensitivity and driver perceived comfort. times / minute and The frequency limit of the massage actuator is times / minute, which is lower than... The massage movements are difficult for the driver to perceive effectively, which is higher than The actuator may generate resonance or noise, which is determined by the physical characteristics of the actuator.

[0063] (3) Waveform switching and part rotation: like Figure 3 As shown, when continuous Response after three (e.g., three) parameter adjustments (including intensity and frequency adjustments) Still below the effective alert threshold When this occurs, it indicates that the current combination of waveform and location can no longer achieve the desired effect through parameter fine-tuning, triggering waveform switching and location rotation to break the driver's sensory adaptation with a completely new stimulation mode.

[0064] The waveform switching sequence is: sine wave (vibration mode), trapezoidal wave (waveform pressing mode), and pulse wave (tapping mode). This sequence follows a gradual principle from gentle to strong, matching the natural process of increasing fatigue. When switching back from the pulse wave is necessary (i.e., the strongest stimulation mode has been reached but is still ineffective), a multi-system coordinated reminder is triggered.

[0065] The body part rotation strategy is: in the waist area Back Shoulders Cyclic switching between these methods, or using a combined activation strategy (such as activating the waist and shoulders simultaneously, denoted as...). The core purpose of site rotation is to avoid localized sensory adaptation and discomfort caused by prolonged stimulation of a single site. A site usage history queue is maintained, prioritizing the selection of recently unactivated sites for switching. Parameters are reset after switching: after waveform or site switching, the intensity and frequency parameters are reset to the initial recommended values ​​corresponding to that waveform or site, avoiding overstimulation caused by inheriting extreme parameter values ​​from the previous combination.

[0066] The above The value of is determined in the following way: This is an empirical value, meaning that the waveform / location should only be switched after the same parameter combination has been tried three times without success. This strikes a balance between giving the parameter adjustment sufficient time to converge and breaking the sensory adaptation in a timely manner.

[0067] (4) Sensory adaptation inhibition mechanism: like Figure 4 As shown, to prevent drivers from developing sensory adaptation to massage stimuli (a physiological phenomenon where the sensory system's response to a persistent stimulus gradually weakens), a habituation inhibitor is introduced. This is used to quantify the degree of sensory adaptation to the current parameter combination. First, it reads the number of consecutive uses of the current parameter combination. Maintain the current massage parameter combination Continuous use counter Each time a massage reminder is triggered, if the parameter combination is the same as the previous one (allowing intensity and frequency to be within range), (fine-tuning within the range), then Add 1; if the parameter combination changes significantly (waveform or part switching, or intensity / frequency change exceeding 10%), then Reset to 1.

[0068] The formula for calculating the habituation inhibition factor is: ; in, The sensory adaptation attenuation coefficient (e.g., 0.3) controls the rate at which sensory adaptation increases with the number of uses; The value range is (0,1], when (When using this parameter combination for the first time) They felt their level of adaptation was low; when (When used 5 times consecutively) The level of adaptation is relatively high, and the reminder effect has significantly decreased.

[0069] Based on habituation inhibitory factors Dynamic compensation is applied to the actual execution intensity, and the compensation formula is as follows: ; in, The current intensity value. For actual enforcement intensity, To control the magnitude of the compensation, a sensory adaptation compensation coefficient (e.g., 0.5) is used. When When the value is close to 1, the level of adaptation is low. When the value is close to 0, the compensation item is close to 0, and the actual execution intensity is approximately equal to the set intensity; when When the temperature drops (the degree of adaptation increases). Increase the compensation item to automatically boost the actual execution intensity to compensate for the effect attenuation caused by sensory adaptation. After the calculation is completed, it must be... Limiting measures ensure that the actual enforcement intensity never exceeds the safety limit. .

[0070] like Figure 4 As shown, when the habituation inhibitor... Below the preset habituation threshold When the value is 0.4, the massage waveform switching or massage area rotation is triggered: the waveform switching (vibration mode, pressing mode, patting mode) is executed in sequence, the area rotation (waist, back, shoulder cyclic switching) is executed in sequence, the continuous use counter is reset, the intensity and frequency parameters are reset to the initial recommended values ​​of the new waveform, and then the massage is performed according to the new parameters.

[0071] The above , , The value of is determined in the following way: Based on empirical data on human tactile sensory adaptation. The higher the value, the faster the adaptation seems to occur; The subjective comfort test was used to calibrate the strength compensation so that it could effectively counteract the effect attenuation caused by habituation, without causing driver discomfort. correspond The system considers the user to have reached a level of familiarity after approximately three uses; that is, if the effect is still unsatisfactory after using the same parameter combination three times consecutively, a switch will be triggered. All of the above parameters are stored as factory defaults in the vehicle's memory, and can also be customized through the vehicle's infotainment system interface.

[0072] In addition, such as Figure 4 As shown, if the current massage parameter combination is used consecutively for a preset threshold number of times, a change in the massage waveform or a rotation of the massage area will be triggered. Through the triple mechanism of waveform switching, area rotation, and intensity compensation, the continuous decay of the reminder effect during long-distance driving is effectively suppressed.

[0073] (5) Multi-system linkage and collaborative reminders: like Figure 1 As shown, when the response degree Less than 0, or response rate after multiple consecutive massage reminders. Still below the effective alert threshold If the seat massage reminder is deemed ineffective, an upgrade linkage mechanism is automatically triggered, initiating multi-system collaborative intervention: Linked air conditioning system: Automatically lowers the interior temperature (e.g., by 2-3 degrees Celsius) or turns on the blower function to enhance the driver's alertness through the stimulation of cold air; Linked audio system: Plays high-volume alert sounds (such as beeps) or voice announcements (such as "Please take a break, it is recommended to stop as soon as possible") to enhance auditory warnings; Linked navigation system: Proactively pushes the location information and navigation routes of nearby service areas, parking lots or rest areas to the vehicle screen to guide drivers to stop and rest in a timely manner.

[0074] The duration of multi-system linkage is dynamically adjusted according to the driver's condition until the driver's fatigue disappears or the vehicle stops moving.

[0075] (6) Safety constraints: To ensure driving safety and driver comfort during massage parameter adjustment, the following safety constraints are imposed on parameter adjustment: Single adjustment range limit: To avoid sudden and intense stimulation or discomfort to the driver caused by abrupt parameter changes, the change in each parameter adjustment is limited to no more than 30% of the current value. Specifically, for intensity parameters, For frequency parameters, This constraint ensures the smoothness and gradualness of parameter adjustment.

[0076] Speed-related intensity limit constraint: Considering the higher precision required for vehicle control at high speeds, excessive massage stimulation may distract the driver or affect operational stability. Therefore, the upper limit of massage intensity is dynamically adjusted based on the current vehicle speed. The specific rule is: when the vehicle speed... hour, (Maximum strength allowed); when hour, (The upper limit is reduced to 90%); when hour, (The upper limit is reduced to 80%). This constraint prioritizes safety at high speeds while ensuring the effectiveness of the reminder.

[0077] Cooling-off time mechanism: To ensure sufficient time to evaluate the effect of the previous parameter adjustment before making the next adjustment, and to avoid oscillations and driver inconvenience caused by excessively frequent adjustments, a minimum cooling-off time is set between two consecutive parameter adjustments. (e.g., 15 seconds), meaning that after each parameter adjustment, wait at least... Only then can the next adjustment be made.

[0078] Driver manual intervention priority: If the driver manually adjusts the seat massage parameters during the massage reminder (such as adjusting the intensity or mode through the vehicle's infotainment system), the adaptive adjustment algorithm is immediately paused, and the driver's manual settings are recorded as a new reference baseline. The adaptive adjustment process restarts the next time fatigue detection is triggered, starting from the driver's manual settings.

[0079] Through the aforementioned dual adaptive adjustment of intensity and frequency, waveform switching and body part rotation, sensory adaptation inhibition mechanism, multi-system linkage and collaborative reminders, and multiple safety constraints, personalized automatic optimization of massage parameters can be achieved to address individual sensitivity differences among different drivers and dynamic response changes of the same driver at different times, while ensuring the safety and controllability of the adjustment process.

[0080] Example 2 like Figure 5 This embodiment provides a driver reminder system based on seat massage. The system is organized according to a three-layer architecture of hardware perception layer, core processing layer and execution layer. It includes a fatigue detection module, a scoring calculation module, a grading judgment module, a massage control module, a responsiveness evaluation module, a parameter adjustment module and a linkage control module. Each module is connected in sequence and works together to form a complete closed loop from fatigue detection to parameter adaptive adjustment.

[0081] At the hardware perception layer, the fatigue detection module includes an in-vehicle camera (preferably an infrared camera) and a steering wheel angle sensor to acquire fatigue feature data from at least two different modalities. The in-vehicle camera captures the driver's facial expressions, eye state (eyelid opening and closing, blinking frequency), and head posture (nodding motion). The steering wheel angle sensor acquires data on changes in the driver's grip strength and steering operation (steering wheel angle change rate, used to detect abnormal operations such as swerving and unstable steering). The fatigue detection module also includes a timing unit to record continuous driving time. These multi-source sensors work together to provide comprehensive data support for accurate identification of fatigue states. The fatigue detection module outputs the acquired multi-modal fatigue feature data to the scoring calculation module. This system does not rely on a single instance of fatigue behavior for triggering; instead, it accumulates scores for multiple features, assigning different weights to each feature based on its fatigue indication intensity. An alert is only triggered when the accumulated score exceeds a preset threshold, effectively reducing the false alarm rate.

[0082] In the core processing layer, the input of the scoring calculation module is connected to the output of the fatigue detection module to receive multimodal fatigue feature data output by the fatigue detection module. This module contains multiple quantization scoring units corresponding to each modality feature: a visual feature quantization unit weighted and synthesized a visual fatigue feature score based on the proportion of eyelid closure duration, yawning frequency, and nodding frequency in facial images; a steering wheel behavior quantization unit weighted and synthesized a steering wheel behavior feature score based on the standard deviation of steering wheel angle and the rate of change in grip strength; a short-term cumulative quantization unit statistically analyzes and normalizes the occurrence frequency of each fatigue feature based on a sliding time window to obtain a short-term fatigue cumulative score; and a duration correction unit calculates a duration correction factor based on the ratio of continuous driving time to a preset maximum safe driving time threshold. This module weights and fuses the above four scores according to preset weights to obtain a comprehensive fatigue score, and outputs the comprehensive fatigue score to the grading judgment module. The values ​​of each quantization score and the comprehensive fatigue score are all within the range [0,1].

[0083] The grading and determination module, whose input is connected to the output of the scoring calculation module, receives the comprehensive fatigue score output by the scoring calculation module. This module internally stores a first threshold, a second threshold, and a third threshold, with the three thresholds increasing sequentially and all falling within the range [0,1]. The grading and determination module compares the comprehensive fatigue score sequentially with these three thresholds: when the comprehensive fatigue score is lower than the first threshold, the driver is determined to be in a non-fatigued state, and no fatigue level is output; when the comprehensive fatigue score is between the first and second thresholds, it is determined to be level one fatigue; when the comprehensive fatigue score is between the second and third thresholds, it is determined to be level two fatigue; and when the comprehensive fatigue score is higher than the third threshold, it is determined to be level three fatigue. The grading and determination module outputs the determined fatigue level to the massage control module.

[0084] The massage control module, whose input is connected to the output of the grading and determination module, receives the fatigue level output by the grading and determination module. Internally, this module stores a mapping table between fatigue levels and massage parameter vectors. Each fatigue level corresponds to a unique set of massage parameter vectors, which include parameters in five dimensions: massage frequency, massage intensity, massage waveform, massage location, and duration. Based on the received fatigue level, the massage control module retrieves the corresponding massage parameter vector from the mapping table and generates a control command, which is then sent to the seat massage actuator. The seat massage actuator includes a lumbar massage area located at the lower part of the seat back, a back massage area located in the middle of the seat back, and a shoulder massage area located at the upper part of the seat back. Each of the three massage areas has an independently controlled massage unit, which can be activated individually or in combination. The massage control module activates a single massage area or multiple massage areas in combination by assigning control signals to different massage units.

[0085] The responsiveness evaluation module has its first input connected to the output of the massage control module to acquire the trigger time information of the massage control; its second input is connected to the output of the scoring calculation module to acquire the comprehensive fatigue score before and after the massage trigger. The responsiveness evaluation module records the comprehensive fatigue score at the massage trigger time as a baseline score when the massage control module sends the control command; after the control command is sent, a timer starts, and after a preset response evaluation time window, the comprehensive fatigue score after the massage trigger is acquired again through the scoring calculation module; then, the responsiveness of this massage reminder is calculated according to the responsiveness formula. The responsiveness characterizes the relative decrease in the comprehensive fatigue score caused by the massage reminder; a larger value indicates a better reminder effect, a positive value indicates a decrease in fatigue, a zero value indicates no change, and a negative value indicates a further increase in fatigue. The responsiveness evaluation module outputs the calculated responsiveness to the parameter adjustment module.

[0086] like Figure 6 As shown, the input of the parameter adjustment module is connected to the output of the responsiveness evaluation module, and is used to receive the responsiveness output by the responsiveness evaluation module. The parameter adjustment module includes an intensity adjustment unit, a frequency adjustment unit, and a habituation suppression unit. The three units work sequentially according to the coordination relationship of the intensity adjustment unit taking priority, and feed back the adjusted massage parameter vector (frequency, intensity, waveform, location, duration) to the massage control module. The intensity adjustment unit internally stores preset effective reminder thresholds, target improvement rates, minimum safe intensity, and maximum safe intensity. The unit compares the received responsiveness with the effective reminder thresholds: if the responsiveness is greater than or equal to the threshold, the reminder is considered valid, and no adjustment is made; if the responsiveness is lower than the threshold, the unit calculates the deviation between the target improvement rate and the measured responsiveness, determines the intensity adjustment direction based on the sign of the deviation (increase intensity for a positive deviation, decrease intensity for a negative deviation), and determines the intensity adjustment amount based on the absolute value of the deviation. It then increases or decreases the current massage intensity by this adjustment amount, finally constraining the adjusted intensity value between the minimum and maximum safe intensity. The intensity adjustment unit feeds back the adjusted intensity value to the massage control module to update the execution parameters for subsequent massage reminders.

[0087] The frequency adjustment unit internally stores the frequency adjustment enable coefficient, frequency adjustment step size, minimum safe frequency, and maximum safe frequency. After the intensity adjustment unit has made a preset number of adjustments (e.g., twice), if the responsiveness still does not reach the effective alert threshold, the frequency adjustment unit is triggered. The frequency adjustment unit calculates the deviation between the target improvement rate and the measured responsiveness, determines the frequency adjustment direction based on the sign of the deviation (increases the frequency when the deviation is positive, and decreases the frequency when the deviation is negative), and then adjusts the current massage frequency by increasing or decreasing it according to the preset frequency adjustment step size, constraining the adjusted frequency value between the minimum and maximum safe frequencies. The frequency adjustment unit feeds back the adjusted frequency value to the massage control module.

[0088] The sensory adaptation inhibition unit monitors and inhibits sensory adaptation to continuous, identical massage stimuli by the driver. Internally, this unit maintains a counter for the number of consecutive uses of the current massage parameter combination (including the massage waveform and massage location), and stores a sensory adaptation decay coefficient, a sensory adaptation compensation coefficient, and a habituation threshold. Each time a massage alert is triggered, the sensory adaptation inhibition unit checks if the current massage parameter combination is the same as the one used in the previous trigger: if so, the number of consecutive uses is incremented by 1; if significant changes have occurred, such as waveform switching or location rotation, the number of consecutive uses is reset to 1. The sensory adaptation inhibition unit calculates the current habituation inhibition factor based on the number of consecutive uses and the sensory adaptation decay coefficient, following an exponential decay model. This factor decreases exponentially with increasing consecutive uses. The sensory adaptation inhibition unit dynamically compensates for the current massage intensity based on the habituation inhibition factor—the compensation magnitude is positively correlated with the decay rate of the habituation inhibition factor, meaning the higher the driver's sensory adaptation to the current stimulus, the greater the intensity compensation—and constrains the compensated intensity value within a preset safe range. When the habituation inhibition factor is lower than the preset habituation threshold, the habituation inhibition unit determines that the current parameter combination has produced significant sensory adaptation and sends a waveform switching command or site rotation command to the massage control module to switch the current massage waveform (in the order of sine wave, trapezoidal wave, and pulse wave) or rotate between different massage areas (waist, back, and shoulders), thereby breaking the sensory adaptation with a completely new stimulation mode.

[0089] Regarding safety constraints, such as Figure 6 As shown, the parameter adjustment module applies the following constraints: single adjustment range limit (not exceeding 30% of the current value), speed-related intensity upper limit constraint (dynamically adjust the intensity upper limit according to vehicle speed), and cooling time limit (shortest interval between adjacent adjustments).

[0090] The linkage control module communicates with the air conditioning system, audio system, and navigation system via in-vehicle Ethernet or CAN bus. The input of the linkage control module is connected to the output of the responsiveness evaluation module to receive responsiveness data. When the responsiveness is less than 0 (i.e., the driver's fatigue level continues to worsen during the massage), or when the responsiveness remains below the effective reminder threshold after multiple consecutive massage reminders, the linkage control module is triggered. It sends a temperature reduction command or a blower activation command to the air conditioning system, a high-volume alert tone playback command or a voice broadcast command (such as "Please take a rest, it is recommended to stop as soon as possible") to the audio system, and a search command and location push command for nearby service areas, parking lots, or rest areas to the navigation system, thus triggering multi-system coordinated intervention. The duration of the multi-system linkage is dynamically adjusted according to the driver's condition until the driver's fatigue characteristics disappear or the vehicle stops moving.

[0091] Between the core processing layer and the execution layer, the massage control module sends control commands to the seat massage actuator, which is located in the execution layer and contains three independent areas: a lumbar massage area, a back massage area, and a shoulder massage area. The parameter adjustment module also imposes the following safety constraints: the amplitude of a single intensity adjustment and the amplitude of a single frequency adjustment shall not exceed 30% of the current parameter value; the time interval between two adjacent parameter adjustments shall not be less than a preset cooling time (e.g., 15 seconds); the upper limit of massage intensity is dynamically adjusted according to the current vehicle speed, with the upper limit decreasing as the speed increases—maximum intensity is allowed when the vehicle speed is not higher than 60 km / h, the upper limit is reduced to 90% of the maximum intensity when the vehicle speed is between 60 km / h and 120 km / h, and the upper limit is reduced to 80% of the maximum intensity when the vehicle speed is higher than 120 km / h. If the driver manually adjusts the seat massage parameters through the vehicle's infotainment system during a massage reminder, the system immediately pauses the adaptive adjustment function of the parameter adjustment module and records the driver's manual settings as a new reference benchmark. The adaptive adjustment process restarts from the driver's manual settings the next time fatigue detection is triggered. The above constraints together ensure that the parameter adjustment process is safe and controllable.

[0092] Through the sequential connection and collaborative operation of the above modules, this system realizes a complete closed-loop control from fatigue detection, grading judgment, massage execution, effect evaluation to parameter adaptive adjustment. It can automatically optimize massage parameters based on the individual sensitivity differences of different drivers and the dynamic response changes of the same driver at different times, while ensuring the safety of the adjustment process and the comfort of the driver.

[0093] Example 3 Embodiment 3 of the present invention provides a vehicle including the driver reminder system based on seat massage described in Embodiment 2.

[0094] During driving, the vehicle monitors the driver's fatigue status in real time through a built-in driver reminder system based on seat massage. When fatigue is detected, the system automatically performs graded massage reminders, responsiveness assessments, adaptive parameter adjustments, and multi-system coordinated interventions, effectively improving driving safety.

[0095] The above description is merely a preferred embodiment of this practice and is not intended to limit the scope of this practice. Various modifications and variations can be made to this practice by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this practice should be included within the protection scope of this practice.

Claims

1. A driver reminder method based on seat massage, characterized in that, include: Obtain fatigue characteristic data for at least two different modes; The fatigue characteristic data are weighted and scored to obtain a comprehensive fatigue score. ; Based on the fatigue comprehensive score The fatigue level is determined by comparing it with a preset threshold. Based on the fatigue level, determine the corresponding seat massage parameters and control the seat massage module to perform massage reminders corresponding to the fatigue level; After the massage reminder is activated, the driver's fatigue characteristic data is reacquired, and the responsiveness is calculated. The responsiveness The fatigue comprehensive score is calculated based on the fatigue comprehensive score at the moment of massage triggering and the fatigue comprehensive score within the preset response evaluation time window after massage triggering; When the responsiveness When the intensity of the massage falls below the preset effective reminder threshold, the massage intensity and / or frequency will be adjusted adaptively.

2. The method according to claim 1, characterized in that, The fatigue feature data is weighted and scored to obtain a comprehensive fatigue score. ,include: Obtain visual fatigue characteristic data, steering wheel behavior characteristic data, short-term fatigue accumulation characteristic data, and continuous driving duration data, and calculate visual fatigue characteristic scores for each. Steering wheel behavior characteristic score Short-term fatigue cumulative score and continuous driving duration correction factor ; The fatigue comprehensive score S is calculated according to the weighted scoring formula: ; in, For real-time sampling time, This refers to the duration of continuous driving during that session. , , , The weighting coefficients are satisfied. , , , , The value range is [0,1].

3. The method according to claim 1, characterized in that, The fatigue level is determined by comparing the comprehensive fatigue score S with a preset threshold, including: When the fatigue comprehensive score When the patient is in the first rating range, it is determined to be level one fatigue. The massage waveform is set to vibration mode and the massage intensity is set to low intensity. When the fatigue comprehensive score When the patient is in the second rating range, it is determined to be the second level of fatigue. The massage waveform is set to the pressing mode and the massage intensity is set to the medium intensity. When the fatigue comprehensive score When the body is in the third rating range, it is judged as level three fatigue. The massage waveform is set to tapping mode, the massage intensity is set to high intensity, and sound prompts or voice warnings are played simultaneously. Among them, the fatigue comprehensive score thresholds corresponding to the first, second, and third score intervals increase sequentially, with the third level of fatigue corresponding to the highest degree of fatigue.

4. The method according to claim 3, characterized in that, When the responsiveness The response rate is less than 0 or after multiple consecutive massage reminders. If the threshold is still below the effective alert threshold, a multi-system coordinated alert is triggered, including: The system can be linked to the air conditioning system to lower the interior temperature or turn on the blower function. The linked audio system plays a high-volume alert tone or voice announcement; The integrated navigation system pushes location information and navigation routes for nearby service areas, parking lots, or rest areas onto the vehicle's infotainment screen.

5. The method according to claim 1, characterized in that, The responsiveness The calculation formula is: ,in, The overall fatigue score is based on the moment the massage is triggered. The response evaluation time window after massage triggering. The overall fatigue score.

6. The method according to claim 5, characterized in that, The massage intensity is adaptively adjusted. This includes: calculating the adjusted massage intensity based on the intensity adaptive adjustment formula. : ; in, For the first Intensity value triggered by the massage For the first The intensity value adjusted upon triggering the massage. For the first The measured response rate of a single massage. For target responsiveness, The step size coefficient is used to adjust the intensity. For the amplitude limiting function, For minimum safety strength, For maximum safety strength.

7. The method according to claim 6, characterized in that, The massage frequency is adaptively adjusted. ,include: When continuous The responsiveness after secondary intensity adjustment The effective reminder threshold has not yet been reached. At that time, the adjusted massage frequency is calculated according to the frequency adaptive adjustment formula. : ; in, This is the frequency adjustment enable coefficient. For frequency adjustment step size, For symbolic functions, For the minimum frequency, This is the maximum frequency.

8. The method according to claim 7, characterized in that, The triggering of massage waveform switching or massage area rotation includes: Calculate the habituation inhibition factor: ; in, This refers to the number of consecutive uses of the current massage parameter combination. This is the habituation attenuation coefficient; based on the habituation inhibition factor... Calculate the actual execution intensity: ; in, The current set intensity value, For actual enforcement intensity, This is the habituation compensation coefficient; when the habituation inhibition factor... Below the preset habituation threshold At the same time, the massage waveforms are switched in the order of sine wave, trapezoidal wave, and pulse wave, or the massage areas are rotated between the waist, back, and shoulders.

9. The method according to claim 1, characterized in that, Also includes: If the number of consecutive uses of the current massage parameter combination reaches a preset threshold, the massage waveform will be switched or the massage area will be rotated.

10. A driver reminder system based on seat massage, characterized in that, include: The fatigue detection module includes an in-vehicle camera and a steering wheel angle sensor, used to acquire fatigue characteristic data of at least two different modes; The scoring calculation module is used to perform weighted scoring calculations on the multimodal fatigue feature data acquired by the fatigue detection module to obtain a comprehensive fatigue score. ; The grading and determination module is used to determine the fatigue comprehensive score. The fatigue level is output by comparing it with a preset threshold. The massage control module is used to determine the massage parameter vector based on the fatigue level. Send control commands to the seat massage actuator; The responsiveness assessment module is used to preset a time window after the massage reminder is activated. Reacquire driver fatigue characteristic data and calculate response. ; The parameter adjustment module includes an intensity adjustment unit, a frequency adjustment unit, and a habituation suppression unit. The intensity adjustment unit is used to adjust the response based on the response. Adjust massage intensity The frequency adjustment unit is used to adjust the frequency according to the response. Adjust massage frequency The habituation inhibition unit is used to calculate the habituation inhibition factor. And compensate for the intensity of the massage.

11. The system according to claim 10, characterized in that, It also includes a linkage control module, which communicates with the air conditioning system, audio system and navigation system via vehicle Ethernet or CAN bus. When the response rate is lower than the effective reminder threshold, the linkage control module triggers multi-system coordinated intervention.

12. The system according to claim 10, characterized in that, The fatigue detection module includes an in-vehicle camera, which is an infrared camera, used to acquire the driver's facial expressions, eye state, and head posture; a steering wheel angle sensor is used to acquire changes in the driver's grip force and steering operation data; the fatigue detection module also includes a timing unit for recording continuous driving time.

13. The system according to claim 10, characterized in that, The seat massage actuator includes a lumbar massage area located at the lower part of the seat back, a back massage area located in the middle of the seat back, and a shoulder massage area located at the upper part of the seat back. The three massage areas can be controlled independently or activated in combination. The massage control module controls the three massage areas to achieve the rotation of massage areas.

14. A vehicle, characterized in that, Includes the driver alert system based on seat massage as described in any one of claims 10 to 13.